My solid container is finally here...cool cnc mpeg

[k|ngp|n]

After many delays, my solid design container has finally arrived!! This is my dream container that I have always wanted. I made absolutely no compromises in making it and it was built with only performance in mind(not to take away from any of the other containers I have used which were all great in their own right). There was flame hardened custom tooling that needed to be made, and re-made, before it could be completed. It may prove difficult to mass produce this particular unit considering the costs that were involved. Along with the custom tooling, the purest copper was used to ensure optimal thermal transfer :/. Some details which are at the heart of the design, are still under NDA and I can't show them just yet. I can say it's a heavy mofo ;). I took some risks, and I hope they pay off. I am going to run it through the ringer with ln2 mostly, and then some DI on both my dothan and FX setups this week coming up, so results coming soon.
I was very lucky to be present and oversee almost the whole production process, and I got a ton of pictures and a killer avi of it being cut on a $250,000 cnc lathe. I will get the vid loaded up some time later tonight and insert it into this post...it's pretty cool.
Here are some misc pics in the meantime:

Production pics:
http://pic8.picturetrail.com/VOL256/1324330/7857526/112221443.jpg
http://pic8.picturetrail.com/VOL256/1324330/7857526/112221439.jpg

Solid block of copper it was cut from:
http://overclockingpin.com/solid%20piece.jpg

Finished product:
http://overclockingpin.com/my%20solid%203%20with%20sig.jpg

Here it is next to my mousepot for scale(of course my mousepot was as shiny as my new container when I first got it, but this is how they look after you use the hell out of them :):
http://pic8.picturetrail.com/VOL256/1324330/7857526/112226252.jpg

short clip of top of the bore getting cut on the cnc lathe:
<a href="http://www.clipshack.com/ClipView.aspx?id=dZzZ2YYYRgnO3qn1Mgmfcw**">
<img src="http://www.clipshack.com/ClipLinkImage.aspx?id=dZzZ2YYYRgnO3qn1Mgmfcw**" border="0">
</a>

[Bloody_Sorcerer]

a work of art.
8 inches tall? sounds short, but what the hell do I know, I've never played with DI/LN2 :)

[Jort]

eyecandy perhaps:)

[harleybro]

Very cool mate! :toast: Lookin foraward to your results. Also what the inside piece is all about. BTW I am now lidless also. ;)

[[XC] moddolicous]

Will it be use with that arc?? Looks really good. Such a good idea that your getting it pattented? Cant wait.

[afireinside]

Looks sick kingpin! Especially the lathe pics :D

[k|ngp|n]

hehe, mine where 9in and to tell ya the truth I think a few inches where not ever needed on my unit :)

those 15in tall units are all show, hehe :)

Thanx for the words bro, the mousepot was a blast and is a tough act to follow, but I am going to try like hell. It will be blood from stone from this point on, but there is still a little more to be had there...especially with ln2.
I always just need more. Prolly why japanese have so many diff revisions.....

And yah J knows how I feel about the tall ass ones, not that there isn't anything wrong with that sort of thing ;)
The mass up top does nothing really IMO. Maybe for DI where there is contstant contact the wole way up...but I believe ln2 goes by a whole different set of rules and you have to decide what the purpose of the container is and design it accordingly. "Put it where you need it " might apply better there.

[G H Z]

Ya baby, big CU in da house ;)

Lookin good KP, hope it does what you design is intended. Say how much does Mousepot weigh in at? I'm guessing around 3.8lbs or so. Looks like we'll be seeing some big numbers over the weekend for sure.

Good luck with it and make sure to do lots of pics & results :D

[justwOo]

Was all you changed between yours and the mousepot more pure copper and a thicker walled base?

Or is there something inside special too? :)

[G H Z]

http://www.xtremesystems.org/forums/images/smilies/rotf.gif http://www.xtremesystems.org/forums/images/smilies/rotf.gif http://www.xtremesystems.org/forums/images/smilies/rotf.gif

[IYP]

mmm midgets

[harleybro]

Darnit so the midget is what I am lacking all along lmao.

[k|ngp|n]

He has fitted a miget inside that controls the flow of LN2 with a small bucket which he splashes around on the inside walls as needed, lol

Im sure he did someting inside on that CNC Lathe. Thats a nice mahcine would be a pity if it was just plain internal :)

hehe..:banana::banana::banana::banana:star midgets. That machine was crazy lol...I had never seen anything like that thing. The vertical shaft mill that they made my hold down on was pretty sick too. Fully automated and made the complete holdown from a solid block to finish from all angles in less than 10 minutes as I watched behind the glass.
The inside of the pot is nothing crazy really and pretty simplistic, just not anything I have yet to see on other solid one piece cylindrical containers yet, japs included. I think it is going to kill with the ln2 on full throttle. Not talking this -75c bs with these noobed up fx's, I'm talkin the real f'in deal maximum -190c full overclocks. If it rocks because of the design, I think it's patentable as it will be unique in that respect. If it sucks, then I will just try a different approach. No biggy. The thrill is in the hunt too. To me even a 50mhz gain is striking gold at that level of benching. I'd just rather not show the base before my testing phase is over. I dont want to see someone else profit from my design if it's a good one,thats all. The honor system doesn't cut it anymore.

I will upload the mpeg in abit.

[\Karting_freak]

That looks kickass KP!
/me waiting for the video /me hopes it wont be as large as 50meg +)

[Equinox]

If you wanted to make a cap that would give a half ass seal for this device you could use a carbon fiber o-ring. I think it expands with cold. I know they use titanium tips for carbon rods on CMM's because the expansion contraction between thoughs two materials. How is everyone sure about what material has the best thermal conductivity? Im not sure either. I've only been going on what gives the best thermal shock resistence. I cant see in the ID, did he flat bottom the work piece? What's the wall thickness between proc and LN2? Very sweet man. Really looks good. Looks to be an older Mazak 860 or 660 slant bed machine he did it on.I used to run the LL7's. you can see he used an Allied 135 degree TiN coated spade. with a littl custom grind. very nice man.

[ilkkahy]

If "heart" was made on lathe there arent much options what kind of constructions there might be. I would put my guess to chilly type of central nod.. perhaps some grooves on pipe too. If a drill was used then there might be something unique inside.

[Waus-mod]

nice tube kingpin. Now i want to see the inside :D and how have you done the base?

[k|ngp|n]

If you wanted to make a cap that would give a half ass seal for this device you could use a carbon fiber o-ring. I think it expands with cold. I know they use titanium tips for carbon rods on CMM's because the expansion contraction between thoughs two materials. How is everyone sure about what material has the best thermal conductivity? Im not sure either. I've only been going on what gives the best thermal shock resistence. I cant see in the ID, did he flat bottom the work piece? What's the wall thickness between proc and LN2? Very sweet man. Really looks good. Looks to be an older Mazak 860 or 660 slant bed machine he did it on.I used to run the LL7's. you can see he used an Allied 135 degree TiN coated spade. with a littl custom grind. very nice man.

You know your stuff man. It was mazak 860 and for the copper, I just went with what was recomended by my machinist friend for my application. There were three or four different grades to choose from. The custom tooling that he made to make my base is not visible in my pics becasue it would give away what is down there. I will get that mpeg loaded this morning, I passed out last night when I came home from partying :/

[SlackeR]

You know your stuff man. It was mazak 860 and for the copper, I just went with what was recomended by my machinist friend for my application. There were three or four different grades to choose from. The custom tooling that he made to make my base is not visible in my pics becasue it would give away what is down there. I will get that mpeg loaded this morning, I passed out last night when I came home from partying :/


I know that feeling too well ;)

Nice container :toast:

[k|ngp|n]

V> Isn't It a nice feeling to use something you made and at that use something that breaks WR's with :)

I got a rush of that and even more so when others used my unit and done the same thing :)

It will be cool to use a container that I made for sure, especially if it works like I want it too and I can break some of my own records with it. It will justify the time, effort, frustration, and cost that I put into it.

I added the clip to the top of the thread btw..I had to cut it down alot from 35mb avi file that it was originally, but it still shows the boring part of the process pretty good. Eventually I will put the the high quality version on my site.
Now if my detroit lions win today and I can at least take back my #2 spot SLI later on or maybe even #1, today will be a great day :)

[Waus-mod]

ok.. im hoping for succes kingpin!

[k|ngp|n]

Well, two personal records down already lol.....if it wasnt for this one POS card, I would have #1 SLI back. I knew this slower card was going to hold me back one day. It can still be done with cpu power though I think.
Lions were getting killed so I gave up on them long ago hehe.

Going to try some dothan pi action next. Maybe I will finally get my 18sec with a 770m.

[[XC] moddolicous]

Any screenies?

[n00b 0f l337]

Sweet and sexy. I hope my alumi (if im not to late) will be this sweet.

[k|ngp|n]

Well...I posted these over at ocXtreme, but here ya go....
http://pic8.picturetrail.com/VOL256/1324330/7857526/112336109.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/112336099.jpg

Still not my best settings by far on the ram, so there is more. I just need this damn FX to cooperate with me a little more. I need to open it up a little with the dothan now. I had 30 litres of ln2 sitting here for two + weeks and it was down to about ten when I went out to the garage to get it earlier, so I barely have any left. Maybe enough for some pentium M pi runs with a pci card in.
400 and some odd points from 50k now fellas....I'm feelin it now.

[k|ngp|n]

Man, I have alot of pics from when it was made...
here is another killer one:

http://overclockingpin.com/container%20making.jpg

[Waus-mod]

Thats a pretty one... cnc is awesome for this work

[SlackeR]

Nice indeed. Now go for 50K, fast :D

[the_chad]

nice work k|ngp|n

how long did it take to complete all up?

cheers,

the_chad

[dastt]

wow

thx for the vid

looks awsome, if i can get into a machine shop in the next week or two im going to do a dice tube

[justwOo]

Wow awesome CNC vid :)

How long total worth of cutting did it take in the CNC machine?
Like interior exterior and everything?

[metro.cl]

you need to work that gpu, shamino i much higher you could take him down if you get that kind of clocks.

keep it coming 50k is so close

[Tribal84]

nice tube kingpin. i love this solid container

[k|ngp|n]

Wow awesome CNC vid :)

How long total worth of cutting did it take in the CNC machine?
Like interior exterior and everything?

It took longer than if it was in a production run because we had to make sure that the tool that was designed to make my base, was doing what we wanted...and it wasn't at first so there was alot of trial and error.

If I decide to mass produce these, we can spit them out at around 1 piece every two hours or so :) That turning machine is the most amazing thing I swear.

[k|ngp|n]

Spent most of the day prepping this p4c800-e I got on ebay for 80.00 bucks :) Full mods for making a dothan scream, mounting, and condensation prepping for something close to -200c takes awhile....that and I am a slacker.
Got to run a few quick benches before the ln2 was all gone, and early on it's looking really good for xtreme ln2 stuff. I manged to get almost -189c loaded temps with about 1.65v on the processor. These are the best temps I have ever gotten on any dothans I've tested under load. The container started hissing loudly as the ln2 boiled off when the temps got below -165c..it was crazy. It must have something to do with the way the base is designed.
Didn't take long to get the chip up close to 3700mhz after seeing what the board could do. What sucks is I couldn't get this mobo to run some of the same ram settings at high fsb as my other good p4p, so I may break it down and hook that other board up for the rest of testing. Dram lat par 11 and 200fsb boot = huge jumps in scores for pi if your board can do it. With both of those settings, I would have had mid 19 sec with this 770 easy. 20.25 sec for 1m was my previous best for 770m but on p4p800.
Also not so sure about this chip, it needs more voltage than my other one did at same clocks, but I did mange to get it 1m stable at nearly 3450mhz with about -120c and 1.6v which is close to 780m performance. 780's like lots of voltage, and so does this one it seems. It should do well at 1.75v+.

Here we go again....
http://pic8.picturetrail.com/VOL256/1324330/7857526/112491262.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/112491267.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/112491273.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/112491258.jpg

This score could be much better with the setings I mentioned above. Something like .6 to .8sec total drop in score.

[harleybro]

Very cool lookin forward to other results! Any way to redo the first pic w/o the skull on the bottom? lol :toast:

[.sentinel]

That is sweet man.

[[XC] DragonOrta]

Sweet Trail of Dead poster man. And awesome initial results. That tube sounds kickin.

[the_chad]

nice work, awesome tube once again, and i like the skull :D

[G H Z]

That hissing sounds wild, must be very quick at transfer. It looks like you have a winner there KP.

[chilly1]

hehe..:banana::banana::banana::banana:star midgets. That machine was crazy lol...I had never seen anything like that thing. The vertical shaft mill that they made my hold down on was pretty sick too. Fully automated and made the complete holdown from a solid block to finish from all angles in less than 10 minutes as I watched behind the glass.
The inside of the pot is nothing crazy really and pretty simplistic, just not anything I have yet to see on other solid one piece cylindrical containers yet, japs included. I think it is going to kill with the ln2 on full throttle. Not talking this -75c bs with these noobed up fx's, I'm talkin the real f'in deal maximum -190c full overclocks. If it rocks because of the design, I think it's patentable as it will be unique in that respect. If it sucks, then I will just try a different approach. No biggy. The thrill is in the hunt too. To me even a 50mhz gain is striking gold at that level of benching. I'd just rather not show the base before my testing phase is over. I dont want to see someone else profit from my design if it's a good one,thats all. The honor system doesn't cut it anymore.

I will upload the mpeg in abit.

PAtent?? O well as I said before get it published and its yours forever patent it and its gone as soon as someone sees it...
What I mean is this.. When ever youcome up with some design make sure to get the drawings notarized.
send then to youe self and publically display the pics..
THis will show you Own it..
THe other issue about Ln2 containers is what you are doing,, the action performed and how it is done. So Not patentable..
Copywrite the look and feel.... YOu copy a design so someone comes up to you and says did you build /sell/ that to so and so.. then that is enforcable...

[k|ngp|n]

Thanx for the valuable info reggie. I have already learned some of this on my own from meeting with others. I'll show the design as soon as I am done testing it's effectiveness. No biggy.

[speed bump]

Nice design man it must feel good to know that you probably have the "coolest" thing around. Hopefully you can throw some volts at that CPU and get a nice 4ghz superpi run in.

[mrlobber]

-1...8...9... Celsius loaded. Absolutely awesome. :worship:

[Eldonko]

Looks great k|ngp|n, awesome work. :woot:

[Waus-mod]

Doesnt this tube provide you the spot for the arc? i dont see the arc :D

[[XC] moddolicous]

Doesnt this tube provide you the spot for the arc? i dont see the arc :D
Was wondering the same thing. Nice results Kingpin. When are we gonna see 18's? Also, are u gonna try another run and get 50k?? Cant wait. HURRY UP!

[the_chad]

Kingpin are you planning on selling these containers once your happy with the outcome?

[k|ngp|n]

Thanx, it's actually a QOTSA poster form when they toured with them, but they're cool too :)
http://pic8.picturetrail.com/VOL256/1324330/7857526/112639790.jpg

Been really busy with other commitments atm, so i get in testing when I can. Did a little bit of pentium M stuff, but I have to work out some p4c/p4p mobo issues first, before I can max either of my chips out. I may runs some single card 3d action on the p4gd1 in the meantime. I want to work the container through a whole range of temps and loads. I manged to get both my 770m and the 780es stable at around 3400mhz with DI temps. These clocks are good for those temps.
770m 206x16 -68c
http://pic8.picturetrail.com/VOL256/1324330/7857526/112643543.jpg
780m 201x17 -77c
http://pic8.picturetrail.com/VOL256/1324330/7857526/112643125.jpg
Much more to come there of course and sorry for the lousy pics.

I did some quick fx57 2d stuff today, and it did great. I matched my best 8m clocks(mhz) to date with DI temps easily. I also got some runs in 1m to pass at 256x15 which I think is around 3838mhz or something. Thats my best 1m mhz too.
http://pic8.picturetrail.com/VOL256/1324330/7857526/112637800.jpg
http://pic8.picturetrail.com/VOL256/1324330/7857526/112637796.jpg
Focus is on the container right now, better multi and ram settings will come later for that stuff. As far as ln2 goes, benching with it has been a dream. My temps barley swing at all under load, even with the fx at 3.8ghz and 1.82v bios.
http://pic8.picturetrail.com/VOL256/1324330/7857526/112637793.jpg
More updates soon, incuding arc stuff :)

[justwOo]

Edit out the skull :)
lol

Im gonna go to san fransisco(store) or something and get something trippy like your skull D:
It's cool

[harleybro]

Nice results so far can't wait for some 3D. :toast:

[[XC] DragonOrta]

off topic, Did you make that skull? because it's the coolest thing I've ever seen

[k|ngp|n]

off topic, Did you make that skull? because it's the coolest thing I've ever seen
No I bought it. It's made of some kind of resin.



the skulls in the way of the internals, hehe J/K looks killer nice temps also :)
that the Replacment 780 dothan?

No. It's that same 780es I made all the 3d runs with. The newer cpuz reports it correctly now. I would love to get my hands on a retail 780m.

[pythagoras]

The tube sings :slobber: Sounds like you have tuning forks int the base, the expansion rates even if they where only a few millimeters long and relatively thin would make them vibrate, so I'm thinking lots of thin pins??

Regards

John.

[k|ngp|n]

The tube sings :slobber: Sounds like you have tuning forks int the base, the expansion rates even if they where only a few millimeters long and relatively thin would make them vibrate, so I'm thinking lots of thin pins??

Regards

John.

The hissing happens between -165c and -190c. Even with the dothan, if I let the container go dry for abit and then hit it with some ln2, it hisses like mad. Louder than anything I have ever used for ln2, it's pretty wild. I can keep my temps at any temperature I want from 0c to -190c with ease.
I am almost done testing the design's effectiveness, and I will show it soon. I am confident now that this thing flat out rocks with ln2 for controlling temps, as it was intended and it does equally as well with DI too.

[k|ngp|n]

My "Ice Dragon" is making this fx scream..total control..how far will she go?

http://pic8.picturetrail.com/VOL256/1324330/7857526/112899923.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/112899928.jpg

[bachus_anonym]

K|ngp|n,

If you ever run 32M with that setup then you will be the most versatile and complete top bencher on this planet:
AC/WC/Phase/DI/LN2 and 1M/8M/32M/Pifast/Aquamark/3DMark :shock:
After you're done reclaiming all 3D spots, finish off with some 32M :) That would be just surreal :D

BTW, sweet container! Waiting for more pics :clap:

[harleybro]

WTG very nice results! :toast: Now it's time for you to hunt for a new card to make up for the lacking one. :)

[k|ngp|n]

Thanx J & HB


K|ngp|n,

If you ever run 32M with that setup then you will be the most versatile and complete top bencher on this planet:
AC/WC/Phase/DI/LN2 and 1M/8M/32M/Pifast/Aquamark/3DMark :shock:
After you're done reclaiming all 3D spots, finish off with some 32M :) That would be just surreal :D

BTW, sweet container! Waiting for more pics :clap:

Thank you bach, I worked hard on the container, and it is paying off.
#1 SLI is within reach, even though Shamino is making nearly impossible.
heh..as for 32m. I have kinda stayed away from it for fear of losing hardware, but maybe I will give it a go when I am done with 3d and some updated 1m and 8m. Didnt OPB lose an fx during a 32m run?

[harleybro]

heh..as for 32m. I have kinda stayed away from it for fear of losing hardware, but maybe I will give it a go when I am done with 3d and some updated 1m and 8m. Didnt OPB lose an fx during a 32m run?
Don't understand why 32m would cause you to loose it. Wasn't his w/ the hole blown in the socket? Must have been due to not enough dielectric grease or too much voltage IMO. :) Don't think it'll die if evrything is set up right. Although I am NOT one to be commenting on condensation problems lol. ;)

Edit: NM was thinkin of the 55 w/ the hole in the socket.

[k|ngp|n]

Don't understand why 32m would cause you to loose it. Wasn't his w/ the hole blown in the socket? Must have been due to not enough dielectric grease or too much voltage IMO. :) Don't think it'll die if evrything is set up right. Although I am NOT one to be commenting on condensation problems lol. ;)

Edit: NM was thinkin of the 55 w/ the hole in the socket.

Maybe I said it wrong dude :)
It is more of a question of "do I really want to burn up a good piece of hardware running a benchmark I am really not that interested in"?

But for you die hard 32m guys, I will run it when I wrap up my fx/sli benching in a day or two here, hows that? Whats the 32m record btw?

[k|ngp|n]

denny at 19m 22s
http://www.xtremesystems.org/forums/showthread.php?t=74368&highlight=denny
then
onepage at 19m 29s
http://holicho.lib.net/pi3355/pi3355.htm

Thanx man :) OPB and denny both have got some serious mem speed working for them there on that one.

I remeber now, my chip that I got from you is the one ranked at #7 there isn't it ;)?

[JacobHansen]

Killer FX speeds..
Nice Going KP!

[\Karting_freak]

KP
common, show us how it looks inside!
or give some general info about it +)

[speed bump]

^^^^if you had the best bench cooling device ever built do you think you would share?

kingpin you and shamino should hook up i bet with your cpu overclock skills and his GPU overlclocking skills you guys would take down 18k in 05.:)

[Jort]

he won't, he put too mutch work into it.

remember asetek and baker block;)

[harleybro]

I remeber now, my chip that I got from you is the one ranked at #7 there isn't it ?
Wait where's this ranking table? :confused: Understand whatcha mean with the 32m. Really lookin forward to what you pull off in 1M and 8M. :)

[k|ngp|n]

Wait where's this ranking table? :confused: Understand whatcha mean with the 32m. Really lookin forward to what you pull off in 1M and 8M. :)

http://holicho.lib.net/pi3355/pi3355.htm

There harley. Jason posted it above.

If i can get the cpu working right at 3840 with the 14x at these timings and bus speed that I ran here below, the scores will be pretty nice :)

ABout the best im getting out of these mushkins:
http://pic8.picturetrail.com/VOL256/1324330/8082894/113132450.jpg
WHere I am atm with cpu power:
http://pic8.picturetrail.com/VOL256/1324330/8082894/113144112.jpg
16x isnt necessary for these cpu mhz btw

[harleybro]

Aha thanks KP. :toast: Lookin good so far if you get those clocks w/ the 14 multi it will be killer. I may give another go soon but lookin forward to gettin bk into the 3d world lol. :)

[Bloody_Sorcerer]

he won't, he put too mutch work into it.

remember asetek and baker block;)
well... for some reason I doubt DI/LN2 will ever go even remotely mainstream, but i understand what you're saying. the xtreme cooling arena is just much too cutthroat these days.

[the_chad]

awesome results, good luck with your aim to hit 3840 with the 14x at those timings and bus speed, i recon you can do it :D

[k|ngp|n]

Time for some xtreme bling bling. We saw how it kicked the fx's ass at di temps, now we will see the true power of my container and what it was actually designed for. It is VERY arc friendly.

http://pic8.picturetrail.com/VOL256/1324330/7857526/113653376.jpg

780es at almost -190c idle in windows at 3900mhz+ with 1.85v on the processor!! Again these are the best idle temps I have ever gotten, and a personal record. Especially considering where the processor clocks/volts are.
http://pic8.picturetrail.com/VOL256/1324330/7857526/113653385.jpg

This shot was taken after pouring about a quarter cup of ln2 in the pot with container base temps at -165c. In almost a blink of an eye after pouring, my temps jump back to -189 and the container roars loudly. It's incredible.
http://pic8.picturetrail.com/VOL256/1324330/7857526/113653382.jpg

These pi times can be much better, and hopefully will be with some work on this worn out p4p800 I am using for testing. This thread is about the container and how it performs with different chips and loads first and foremost, everything else is secondary. I used this motherboard because I am able to give the chip over 1.85v without issue from ovp/ocp.
With the usual tweaks of dram lat par 11, 200fsb boot, and maxmem 104 this 1m is prolly around
19.5 - 19.6 seconds. Definitely not at the chips limit mhz wise either.
Picture was snapped right as benchmark ended:
http://pic8.picturetrail.com/VOL256/1324330/7857526/113653396.jpg

How about .2 -.3c load temps rise during 8m with a 780es at close to 1.9v and 3900mhz+? :slobber:
I was so floored when I ran this bench, I had to run it three more times to make sure it was for real, and it did it flawlessly all three times:
http://pic8.picturetrail.com/VOL256/1324330/7857526/113653392.jpg
http://pic8.picturetrail.com/VOL256/1324330/7857526/113653942.jpg

Again, the time isn't all that really and could be hugely better with the above mentioned tweaks and some tighter timings. I tweaked nothing here. This is just to show what the container can do with the pentium M close to it's limits mhz/volts wise. I want to see how far I can push the chip mhz first, then I will start tweaking and tuning to get the times down.

http://pic8.picturetrail.com/VOL256/1324330/7857526/113654657.jpg
I'll be posting more pentium M stuff soon...

[SlackeR]

That container is unreal :)

[.sentinel]

Wow Kingpin remeber when we couldn't bench FX'es towards 4 ghz and now it is dothan.

[metro.cl]

awsome mate.

You have to begin benching for WR now

[Vegitto]

No more work on this one?

[G H Z]

Damn KP,

That container really looks to be all you wanted and more. Awsome work on that for sure. That post pour vapor cloud is not normal for any other container?

[k|ngp|n]

No more work on this one?

My prototype and design testing is done, and the unit that you see here will be the final production unit that I am making if I decide to mass produce it.
The cpu units could be available in about a week or so, without pre-order.
As far as my own personal benching, I have to wait for the weekends mostly because I have been working alot and I just dont have the time. I am going to give 50k 01' one more shot this weekend, and some dothan pi stuff.
Honestly, I gave up on 3d when I saw the new ATI cards coming out but...now it looks like I have some more time. I thought for sure someone would have posted over 50k already lol.... guess not.


Damn KP,

That container really looks to be all you wanted and more. Awesome work on that for sure. That post pour vapor cloud is not normal for any other container?

Boiling off action and vapor clouds of ln2 is normal for any container G, and usually when the container gets closer to the temp of the ln2 iteslf, it calms down and gets quiet. For this one, the ln2 gets crazy and loud and really comes to life once you get BELOW -160. Going from -160 to -190c, it roars when you pour it.
It gets my heart going every time I swear :)
I am thinking it is due the increased surface area near the base and the way the base is designed. UNlike other containers I used, this container operates at it's xtreme max cooling capacity when a certain ammount of ln2 is in it, or if using DI it is full of crushed DI and acetone. ALL other containers I have run for ln2, actually performed worse with over a cetain ammount of ln2 in them. Too much ln2 and your evap temps go up.....

[Waus-mod]

Kingpin.. how much insulation do you use on the back of the motherboard between the arc? how thick?

[speed bump]

Come on 3dmark01 you can do it

[k|ngp|n]

Damn KP,

That container really looks to be all you wanted and more. Awesome work on that for sure.

I did pretty good on MY FIRST REVISION huh?
I figured my results would'nt be too hot since this is only MY FIRST REVISION of a cpu container ever.
You would think that 3840mhz on an fx at 1.83v and 3814mhz for 3d and -189c loaded with the dothan benching at 3.9ghz+1.9v could only be the work of a true MASTER, not someone like myself who only just built his FIRST REVISION :)
Guess I got lucky...

I got a shiny new retail dothan 780m sitting here G, so the real issue is are these 780 retail chips fsb limited like the 780es's?
http://pic8.picturetrail.com/VOL256/1324330/8242801/114352185.jpg
I hope it's a good clocker It's RH80536 sl7vb Let's see what the chip can do with crushed DI and acetone first with dragon/p4p800-e/ct-479/ 2x256 buffalo tech bh5.

[n00b 0f l337]

Not bad not bad...

[metro.cl]

good luck mate, you are the best bencher right now (maybe shamino is up there with you)

[illmatik]

awsome mate.

You have to begin benching for WR now

Good Point. What is the dothan WR now? This is definately the fastest dothan I've seen so far.. wow is all I can say, incredible!

[G H Z]

I got a shiny new retail dothan 780m sitting here G, so the real issue is are these 780 retail chips fsb limited like the 780es's?


That chip should fly, hopin for some fast ram speeds. I'd really like to see you try a 670 & single card X1800XT or XL. I'm thinkin I might try it myself http://www.xtremesystems.org/forums/images/icons/icon12.gif

I also have a question. I tried DI on my FX for the first time last week and it sucked. Nothing stable even @ low ass 3.5Ghz. Tried every voltage under the sun. I noticed you said a too full container can perform worse and I'm wondering if that was my problem cuz I didn't have temp monitoring on the container. Too me it just seems like you want the container 1/2 full of acetone :confused: Or is it better to have it only near the bottom and let the DI fall into it?

Btw this FX does 3668 3D stable @ 1.69v on Baker unit & the container had perfect contact.

If you have any ideas I'd appreciate it http://www.xtremesystems.org/forums/images/icons/icon7.gif

[k|ngp|n]

That chip should fly, hopin for some fast ram speeds. I'd really like to see you try a 670 & single card X1800XT or XL. I'm thinkin I might try it myself http://www.xtremesystems.org/forums/images/icons/icon12.gif

I also have a question. I tried DI on my FX for the first time last week and it sucked. Nothing stable even @ low ass 3.5Ghz. Tried every voltage under the sun. I noticed you said a too full container can perform worse and I'm wondering if that was my problem cuz I didn't have temp monitoring on the container. Too me it just seems like you want the container 1/2 full of acetone :confused: Or is it better to have it only near the bottom and let the DI fall into it?

Btw this FX does 3668 3D stable @ 1.69v on Baker unit & the container had perfect contact.

If you have any ideas I'd appreciate it http://www.xtremesystems.org/forums/images/icons/icon7.gif

1/2 full is way too much bro. Here is what ya do for awesome evap temps:
Crush up the DI real good with a hammer or something in like a double bag. Get to almost like it's ice snow and all crushed. Fill the container up with the snow using a funnel or something (I use a scooper) Tamp it down good then pour in ABOUT A SHOT or two at most of acetone. And your good to go. This will have the DI touching all available real estate inside the container and no gaps between chunks. Just refill the crushed ice as it goes down in the container and you wont need to add any more acetone for rest of session.

My first few DI session I used way too much acetone and it was nasty. I can still remember it. I get the chills and have flashbacks when I see those tall ass containers with bubbling and big chunks floating around :)

[babyelf]

Just refill the crushed ice as it goes down in the container and you wont need to add any more acetone for rest of session

so use the crushed powder like dry ice all the time or just the bottom layer before starting?

[harleybro]

Very nice KP lookin forward to new dothan runs. :)

[babyelf]

always use crushed ;) I started that and It works awsome.
Only donwside is if its really humid the crushed ice=more surface area for mostiure to attach to each partical and water make a barrier and Is not good in the mix so try not to let the crushed ice sit with alot of humidity for too long :)

cheers for that :)

will try it next time ;)

so crush it just before i need it?

[G H Z]

1/2 full is way too much bro.

OK I'm inspired to try again, I was trying to convince myself it was cold bug or something ;)

Thanks :toast:

[k|ngp|n]

OK I'm inspired to try again, I was trying to convince myself it was cold bug or something ;)

Thanks :toast:

NP G. Your on the right track with DI now. WHen I was referring to the container being too full and perfroming worse, I was actually referrring to ln2.

ABout this 780retail guys....After some quick testing on DI/acetone, I can say for 100% that these chips are not fsb limited like the 780es's. I was able to hit 3500mhz/250fsb for 1m easily with only -71c evap temps on my pot, and 8m crapped out about 3/4 way thru at 3500.(machine check exception error)...cpu just needed more power. This is with only uwiremod and droop mod..so I am thinking that this prolly a 4ghz chip on max ln2. Imagine what close to 4ghz with high fsb could do with 1m and even 8m :slobber: My 780es could only use 17x after 3330mhz or so just like the japs chips. THat chip maxed out around 3400 with DI and around same volts.
THis 1m is still crappy time as it is on my p4p800-e that can't run 11dram or 200 boot for full pat. Maxmem104 ws not used here either. JUst straight clocks.
http://pic8.picturetrail.com/VOL256/1324330/8242801/114426048.jpg

[k|ngp|n]

Here's the money shot fellas.
Shhh... I wouldnt want anyone to incorporate this into their design because it :banana::banana::banana::banana:ing owns ;)
http://pic8.picturetrail.com/VOL256/1324330/8242801/114561940.jpg

[ilkkahy]

Here's the money shot fellas.
How long pole did you put there? It seems low from the picture but pictures lie often.

No offense but i dont think this desing quite meets the requirements of patents? :rolleyes: Ok in USA you may have different system and you can sue anyone just about anything. I find this ridiculous. For example MM new container has very similar system. Perhaps japanese should try forbit using solid containers now because they were the first to use it? :D In reality you just need to put those in production fast before others.

ps. I am absolutely positively sure i have seen this desing as non solid version like a year ago made by some japs.

All same to me though..

[k|ngp|n]

How long pole did you put there? It seems low from the picture but pictures lie often.

No offense but i dont think this desing quite meets the requirements of patents? :rolleyes: Ok in USA you may have different system and you can sue anyone just about anything. I find this ridiculous. For example MM new container has very similar system. Perhaps japanese should try forbit using solid containers now because they were the first to use it? :D In reality you just need to put those in production fast before others.

ps. I am absolutely positively sure i have seen this desing as non solid version like a year ago made by some japs.

All same to me though..

I can't say how long the pole is becasue I think this is very critical part. The container is full of ln2.
I don't have a patent on it becasue your right it doesnt meet the requirement for patent. It has to be something completly new for that. My specific design is copyrighted, but not patented of course.
I am about 100% positive that this is the only solid one piece container with this design.....and believe when I tell you it WORKS VERY VERY WELL.
I still haven't even decided if I want to mass produce this yet. I have enough material to make about 15 pieces, but I was thinking of just keeping the material and trying all different kinds of designs. I have good access to the cnc.
I just thought that I would show you guys what was down there. Lets see how many containers come out looking like this now :)
I am not trying to get rich lol..I just want to make the best possible cooling device I can.

[Lestat HWL]

Doesn't look bad Vince...The walls look a bit thick...They look around 3/8" thickness, so I'd assume the base is 1/2 or 5/8" thickness. Maybe more after the base peice.

Seems like that'd be a bit too thick and would take awhile to get cooled down...What do you think? I'm sure you looked into this alot before building the tube ;)

[k|ngp|n]

Doesn't look bad Vince...The walls look a bit thick...They look around 3/8" thickness, so I'd assume the base is 1/2 or 5/8" thickness. Maybe more after the base peice.

Seems like that'd be a bit too thick and would take awhile to get cooled down...What do you think? I'm sure you looked into this alot before building the tube ;)

Hey Lestat :)
MY exact measurments on the base and lower wall thickness(below holdown) is a secret, becasue honestly I think that makes or breaks a container.
But you're very right bro, it does take awhile to get to the temps down to max ln2. That was my biggest complaint with the original mouspot...and this one even takes a bit longer because of the added mass.. But like the mouspot, once the temps get where you need them, they stay there. THis is the way of the solid. I was looking for temperature control with ln2 for the fx, and thats why I built it the way I did. I think all solid one piece containers probably behave this way more or less. I was just used to chillys container, where you could get to say -180c very quickly, but not as much capacity for holding temps wihtout keeping the container full of ln2. Mine only needs small splashes and I can actually run the fx with it empty almost.
I combined elements of chillys spiral evap containers that I liked when benching with some elements of the orig mouspot and came up with this frankenstein and I couldnt be happier with the way it turned out. My biggest fear was that the rod was too much and would prevent the base from cooling properly, but it ended up the ultimate in temp control.

[EZClock]

Your results are insane, good job. Any chance for 4ghz Dothan runs? 19 seconds would be possible @ that clock speed id imagine.

[Qkjhfhaiguihfma]

Hey Lestat :)
MY exact measurments on the base and lower wall thickness(below holdown) is a secret, becasue honestly I think that makes or breaks a container.
But you're very right bro, it does take awhile to get to the temps down to max ln2. That was my biggest complaint with the original mouspot...and this one even takes a bit longer because of the added mass.. But like the mouspot, once the temps get where you need them, they stay there. THis is the way of the solid. I was looking for temperature control with ln2 for the fx, and thats why I built it the way I did. I think all solid one piece containers probably behave this way more or less. I was just used to chillys container, where you could get to say -180c very quickly, but not as much capacity for holding temps wihtout keeping the container full of ln2. Mine only needs small splashes and I can actually run the fx with it empty almost.
I combined elements of chillys spiral evap containers that I liked when benching with some elements of the orig mouspot and came up with this frankenstein and I couldnt be happier with the way it turned out. My biggest fear was that the rod was too much and would prevent the base from cooling properly, but it ended up the ultimate in temp control.


careful with those numbers, that's enought info to back into the base thickness etc., or at least give someone a head start. if you really want it to be a secret, i would take out the weight. nice work tho :toast:

edit, you took the weight out :D

[Eldonko]

Hey Lestat :)
MY exact measurments on the base and lower wall thickness(below holdown) is a secret, becasue honestly I think that makes or breaks a container.
But you're very right bro, it does take awhile to get to the temps down to max ln2. That was my biggest complaint with the original mouspot...and this one even takes a bit longer because of the added mass.. But like the mouspot, once the temps get where you need them, they stay there. THis is the way of the solid. I was looking for temperature control with ln2 for the fx, and thats why I built it the way I did. I think all solid one piece containers probably behave this way more or less. I was just used to chillys container, where you could get to say -180c very quickly, but not as much capacity for holding temps wihtout keeping the container full of ln2. Mine only needs small splashes and I can actually run the fx with it empty almost.
I combined elements of chillys spiral evap containers that I liked when benching with some elements of the orig mouspot and came up with this frankenstein and I couldnt be happier with the way it turned out. My biggest fear was that the rod was too much and would prevent the base from cooling properly, but it ended up the ultimate in temp control.
It is aprox 8inches high, 2.5inches wide, and still wieghs 4.6lbs. That should say something about the mass right there.
That is great stuff KP, very innovative, temp controllable LN2 that can be used for an FX is groundbreaking. Sounds tough to controll the temps over a period of time, but I'm sure you can do it. Look forward to your benches.

[G H Z]

OK I know your prolly busy busting 4Ghz PI but I'm sure you can spare a couple of those Mhz to post some screens ;)

So what do you think KP, is a 670 in your future? We need a run @ 50K on ATI :D FIrst guy there will be remembered. Too bad the regulars have to wait for the XT, macci will probably knock that out before anyone else gets a chance.

[k|ngp|n]

OK I know your prolly busy busting 4Ghz PI but I'm sure you can spare a couple of those Mhz to post some screens ;)

So what do you think KP, is a 670 in your future? We need a run @ 50K on ATI :D FIrst guy there will be remembered. Too bad the regulars have to wait for the XT, macci will probably knock that out before anyone else gets a chance.

Heh...I can't go for 18sec until I grab a new p4c800. I am hunting for a few of them in online auctions. I am betting that this chip is gonna do it.

Speaking of 50k, I am runnig SLI later on today for the first time since I got almost 17.2k 05'.... we'll see what happens ;).
Spent all last night trying to break into 21sec using only 15x and I am so damn close. Look what I was able to here with only DI on my container and 1.7v!!:
http://pic8.picturetrail.com/VOL256/1324330/7857526/114794491.jpg
http://pic8.picturetrail.com/VOL256/1324330/7857526/114797098.jpg
THe load temps are incredible with DI at 3825mhz. Start off 8m around -72.5c evap temps and by loop 16, temps were about still below -71.6
At 3800mhz+ during 1m...temps dont even move lol and stay right at -72.5 or so.
http://pic8.picturetrail.com/VOL256/1324330/7857526/114797096.jpg
:slobber: 3837+ with 15x can take the sp1m WR. You have to be able to run insane timings on the ram though. I almost got my mushkins right where I need them now.
Remember G, the last time I ran 01' and got 49.6k and I was only running at 3780mhz with 15x, with no where near these timings on this 1m. I have since run 05' at 3814mhz and there is more I know it. I almost got this thing 100% stable for 8m at 3825 with same settings, so I know it can easily run 01' at about same speeds. I am going to be trying like hell for the next couple days to get 50k for real.....400 points is there, I just need to make it happen. My cards won't go any farther really, so I need to coax the points out of the system.
To tell you the truth I would rather do it with the 7800's at this point because, it is much harder. Anyone can go out and get new hardware to break last gen records..thats the only way some guys can do it ;).
I cant wait to bench with the ati cards...but I am sure 50k will be easy with them and not much of a challenge.

[jan]

great results kingpin :)

[G H Z]

Wow that PI time is killer. Actually 50K single card is what I mean ;)

[k|ngp|n]

GOTTA GIVE THAT CHIP 1.776V ;)
Nice temps M8 on the unit :)


This thing is the wonder chip eh? I should take a close up of the core for ya....I cracked two corners off it the other day, and it still rocks.
And thanx...they really are damn good loaded temps for DI at these mhz.
This cpu can take 1.82v... and actually need close to that for 3d/sli. YOu'd think 8m stability would mean stability for 3d but it isn't the case. THis chip needs more power for 3d.
I was pretty much amazed that it was stable on the DI at that voltage up to 3825, and almost 3840mhz for 1m and nearly 8m. I was about 2 loops away from 21.8sec with 15x I swear.

[T_M]

king, whats the BIOS reported CPU temp when you're are getting those -72C reading with the probe?

[G H Z]

I'll bet he's close to -34° I think BIOS temp is usefull but it's not everything. Different container designs may report similar temps idle in BIOS, but when the OS is loaded and the real action begins the differences can be striking. BIOS rev can play a part too but from what I have seen on NF4 they are very consistent rev to rev.

OK I'm done interjecting, maybe KP can actually answer your question now ;)

[k|ngp|n]

-37 to -38c with 1.7v+ on the chip and -71c base temps.

[T_M]

seems a little high IMO, but im just speculating (my venice at 1.7V+ reads -45C).
does the extra cache add heat?

edit: oh, is that with or without IHS? Mine was without.

[k|ngp|n]

seems a little high IMO, but im just speculating (my venice at 1.7V+ reads -45C).
does the extra cache add heat?

edit: oh, is that with or without IHS? Mine was without.

Fill your container with DI / acetone. Put in a fx57. Set it at 1.55x113%, then clock it to 250x14. Reboot, and go into your bios and tell me what temperature you got compared to your base reading at -71c to -72c.
How do you measure base temps btw? What instrument and positioning of the probe exactly?

[T_M]

I do not measure base temps, they mean nothing to performance IMO.
I also do not own an fx-57 ;) hence why i asked if there should be much of a temp difference between that and a venice both with IHS removed and same volts.

My in-pot temp measured at -66 with my DMM (rated to -20C), and that -45C was windows idle (mbm) and also BIOS idle (DFI)

[harleybro]

I do not measure base temps, they mean nothing to performance IMO.
I also do not own an fx-57 ;) hence why i asked if there should be much of a temp difference between that and a venice both with IHS removed and same volts.

My in-pot temp measured at -66 with my DMM (rated to -20C), and that -45C was windows idle (mbm) and also BIOS idle (DFI)
IMO base temp. measuring is extremely important especially running LN2 on a cold bugged chip. With the 57 I have run, the Opty, and a few SD's I have seen results pretty close to what KP showed. I believe -34c bios or so puts me around -68 - -70 at the base. It pretty much seems to scale accordingly at that range. :)

[speed bump]

Ahh K|ngp|n thanks for finally showing the inside of your container it has helped me out with a quandry I have had about really getting the core cold and maintaining load. Time to start drawing something up.

[k|ngp|n]

I do not measure base temps, they mean nothing to performance IMO.
I also do not own an fx-57 ;) hence why i asked if there should be much of a temp difference between that and a venice both with IHS removed and same volts.

My in-pot temp measured at -66 with my DMM (rated to -20C), and that -45C was windows idle (mbm) and also BIOS idle (DFI)

....
If you do not own an fx57, then why are we talking here? Honestly, what relevant input do you have for my thread? Anyone who knows anything about base temps vs bios temps will disagree with you about "your opinion" I promise.
LOL...your talking about " in pot " temps and telling me evap temps dont matter?????? WTF does inpot temps have to do with anything?
Like I said...get an fx57, run my settings, fill your container up, and post what you get k? If you are running a thin walled container, I can promise you wont be anywhere near my numbers.
I am done with raging.


Ahh K|ngp|n thanks for finally showing the inside of your container it has helped me out with a quandry I have had about really getting the core cold and maintaining load. Time to start drawing something up.

Honestly, It think this design can't be beat for best possible loaded temps.
They are rock stable and the funny thing is this container was orignally desgned for ln2 first, DI second...but it is turning out to be unbelievable for DI. I guess that is what testing is for :)

Now let's get to the real business at hand gentlemen....
http://pic8.picturetrail.com/VOL256/1324330/7857526/115096522.jpg

Less than 225 points from 50k. My ice dragon did this with only DI + acetone :) The capacity on my container is :banana::banana::banana::banana:ing insane. I never thought it could be this easy with DI on this chip. Always needed ln2 before to run these speeds/volts for 3d....Lets see if 50k goes down by sunday night....

[T_M]

So the minute someone has a question for you, you jump on the defensive and think you are too good for me?
righto then..... :stick:

glad GHZ and HB could respond with some decency :clap:

my in pot temp was measured for the hell of it, is that ok with you? only reason i told it to you was to possibly add some info to my weak argument (which you have obviously no intention of actually replying to, just ignoring).

sorry if im not hardcore like you

PS - heres a BIOS shot of a 3500+ Clawhammer (130nm) of mine at 1.85V just for further weight (-36C on a hotter chip than FX-57)
http://img.photobucket.com/albums/v310/T_M/IMG_0354.jpg

[T_M]

I just dont like being told to basically get :banana::banana::banana::banana:ed because i questioned his temperatures, sure i may be heading down the wrong path trying to even compare Venice/Clawcastle temps to double cached 57's but at least tell me why

[k|ngp|n]

So the minute someone has a question for you, you jump on the defensive and think you are too good for me?
righto then..... :stick:

glad GHZ and HB could respond with some decency :clap:

my in pot temp was measured for the hell of it, is that ok with you? only reason i told it to you was to possibly add some info to my weak argument (which you have obviously no intention of actually replying to, just ignoring).

oh and im sure that MM would have something to say about you critiscising his pots

sorry if im not hardcore like you

PS - heres a BIOS shot of a 3500+ Clawhammer (130nm) of mine at 1.85V just for further weight (-36C on a hotter chip than FX-57)
http://img.photobucket.com/albums/v310/T_M/IMG_0354.jpg

Dude..I dont think I am too good for you, I am not on the defensive, and I thought I coverd what it was that you were saying in my resposne. Let me try again..

The only way to compare bios vs base temps is to use the exact same hardware with same settings/clocks/cooling etc. That is what I tried to communicate. Sorry If it sounded harsh...wasn't intended like that.
Base temps are important in lots of different ways. That is my opinion.
I dont know anything about how much heat extra cache adds, so I didn't respond to it. Again, bad comparison.
Enough said.
50k could happens soon fellas......don't ruin the moment. Let's just be happy, k?

kp

[T_M]

well, forgetting totally what base temp you or I may or may not measure, the chip temp in BIOS seems high to me.

[k|ngp|n]

well, forgetting totally what base temp you or I may or may not measure, the chip temp in BIOS seems high to me.

If you had an fx57 clocked at 3500mhz with 1.55vx113% in bios, you would think they were good temps...believe me. 37-38c in bios with these settings and -71c to -72c is pretty good. Event that is just idle temps though and doesnt speak for the whole story. What really matters the most, is what happens to temps under load.

[Qkjhfhaiguihfma]

less debating more 3dmarking!!

you can dooooeeet

[k|ngp|n]

less debating more 3dmarking!!

you can dooooeeet

It's going to be very difficult. This score was not easy at all. My hardware is torched. Cards are running great, but are beyond maxed, core on my fx is chipped in two spots, ram is const erroring out, and the board is pretty worked. My mushkin bh5 is not working at 14x 3750+, or the fx mem controller preventing the ram from working right depending on too low temp or high mhz.

[k|ngp|n]

Still at the top props and congrat's on that for sure :)

Thanx bro. Can never thank you enough for the chip :). I cannot believe what this thing has been through, and is still giving me better numbers as of this mornng. You know I am using that same SLI-DR again now too that I thought died? It just started working again one day, prolly just moisture in the socket and needed to sit for awhile. Always happens like that ;)...

[harleybro]

Thanx bro. Can never thank you enough for the chip :). I cannot believe what this thing has been through, and is still giving me better numbers as of this mornng. You know I am using that same SLI-DR again now too that I thought died? It just started working again one day, prolly just moisture in the socket and needed to sit for awhile. Always happens like that ;)...
Awesome score! Sounds like me throwin the stuff you thnk is dead on the side ony to be resurected at a later date. :D

[zakelwe]

In reference to Venice v Fx-57 the FX-57 uses about 12-15% more power at idle than a Venice at same speed. However, this is at 2800Mhz and at idle.

http://www.lostcircuits.com/cpu/amd_fx57/12.shtml.

I agree with Vince, unless you had exactly the same hardware apart from the tube you cannot really compare one tubes performance to another, just too many variables.

Vince I think being so close to 50k ( with no sleep ? ;) ) is making you a bit cranky though .. cheer up, that's a fantastic score :toast: .. go on, knock down 50k before XT1800 gets it and then Futuremark can retire it before 06 .. :D

Regards

Andy

[k|ngp|n]

In reference to Venice v Fx-57 the FX-57 uses about 12-15% more power at idle than a Venice at same speed. However, this is at 2800Mhz and at idle.

http://www.lostcircuits.com/cpu/amd_fx57/12.shtml.

I agree with Vince, unless you had exactly the same hardware apart from the tube you cannot really compare one tubes performance to another, just too many variables.

Vince I think being so close to 50k ( with no sleep ? ;) ) is making you a bit cranky though .. cheer up, that's a fantastic score :toast: .. go on, knock down 50k before XT1800 gets it and then Futuremark can retire it before 06 .. :D

Regards

Andy

No sleep last night for sure andy. I think I am finally done here with this setup.
My lucky board stopped working again last night in the middle of benching, and I can't seem to get this chip over 3750 or so now with the other two dfi's I have here. Pretty sure it's the chip. I found one of those little caps around the core on the chip partially crushed, and I think it's keeping me from hitting 3800mhz+. I have no idea how I did it. The chip can still rock 8m up to 3750, but higher seems very unstable. It's pretty sad when anything less than 3814mhz isn't enough.
I gave it all I could...with a better 7800gtx to match my good card, 50k would have been done days ago no doubt about it. Guess I pushed my system a little too hard trying to make up for it. :(
Time to retire this SLI rig. I still feel so dissapointed that I couldn't do it, even though the score is tops.

[G H Z]

I still feel so dissapointed that I couldn't do it, even though the score is tops.

Ya it drives you crazy when you only need a few more points :confused2

Nice run KP :up:

[k|ngp|n]

It's cool G....this hardware has been through alot of testing in the last 2 months or so and I am surprised it made it this far.

Here are some parting shot from my last session. This is what I call perfect contact :). These are about the best bios vs evap temps that I got with these settings during my testing with this hardware. Here is something measurable for compare. This was done with DI +acetone on my container. Again, best bios vs evap temps I have seen yet at these settings on the fx57/dfi sli-d latest bios:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115420081.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/115420082.jpg

Here is a nice one with fx57 on a nf4 ultra-d that I modded to sli last night using 623 bios and def clocks and volts. Temps seem to read a tad colder on this mobo:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115420076.jpg

Bios temps vs evap temps basically show your temp delta and what the container is capable of producing for on die evap temps at a given base temp with or without a load.
The key is to have the temps rise as little as possible under load. I have tried to show here that this is best achieved through internal design and working surface, NOT internal volume. All more volume does is lessen refill time.
My dragon is 8 inches high and 2.5 inches wide and can still run 3d and hold -71c to -72c evap temps on fx with container half full of chunks + acetone. This speaks volume about it's ability to hold a load:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115420077.jpg

[k|ngp|n]

Damn he's got alot of acetone in there ;)
I was pretty surprised myself that my container ran that well half empty, Iw as busy pouring the gpu containers, so I didnt even realize it went empty mid-bench.
Opb 32m is quite impressive I have to admit, he's the man for 2d. But, those guys and sham have the best fx-57's so far for sure. I could have done some serious damage with a chip that good. Maybe I should start chip hunting again :)

[k|ngp|n]

SLI 3d is out for the time being.. so time to go back to finish up my full ln2 dothan testing with my container. I have always wanted 18sec 1m with a dothan, and I am now closer than ever. Was playing around tonight on my lousy p4p800 that can't run dram11 or 200 boot for full pat with my original 780es and my dragon pot with ln2 and I am now stable at 3983mhz for 1m :)...soon hopefully 4ghz. I could never get this chip stable over 3950 before.
http://pic8.picturetrail.com/VOL256/1324330/7857526/115689415.jpg
These clocks here with dram11 and 200 boot on my p4c800-e should give me very close to 18sec and a nice 8m time too. I am modding it up for tommorrow.

Got Cpu so far to around 4050, but it was unstable. I am not really into windows stable screenshots, but I took one for the hell of it. 4ghz on a dothan is a feat. I hope to 1m at these speeds closer to 1.9v:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115689416.jpg
BTW--Here is a pic of my 780es that I benched with till 4 in the morning: lol...it looks like someone sanded down all the corners:
http://pic8.picturetrail.com/VOL256/1324330/8242801/113996039.jpg

p4p800 all modded up and totally sucking...board cant bench stable with 200 boot(for full pat) or dram 11 over 225fsb!:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115689417.jpg

frosty:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115689418.jpg

[Waus-mod]

that last shot is awesome kingpin.. i love youre arc.

[harleybro]

Really awesome Dothan run! Impressive scores. 18 1M would be crazy. :eek:

[k|ngp|n]

Getting closer :)...
If it's not one thing it's another.
Having some problems with this p4c and getting the right Volts to the chip for above 3955 1m stable..I'll get it sorted. Also cpuz is making the rig crash out when I open it above 3500mhz?? Maybe corrupted Os.. Never seen that b4, but anything goes with these chips.

almost 1/2 sec away now:
http://pic8.picturetrail.com/VOL256/1324330/7857526/115840818.jpg

[afireinside]

Where the hell did all your pics go man? Friend was looking for them... And you've been at 19s forever now. Do an 18!

[k|ngp|n]

Before this morning, my best 1m pi was 19.89 sec or something like that at 3902mhz with p4p800-e. I did that with this same chip like a month before I started benching sli. It's on my website and hasnt been updated yet. I'll update my best PI time when I get it here soon. All of my best scores and old screenshots are there now. The pi I posted here at 19.6 is my new personal record with any dothan chip :)...I think I may have edged past roger(the stilt) and his prescott, but not sure. This is my lowest time ever, about a .3 second gain from last time I ran. I was able to run 3960mhz+ on the p4gd1, but the pi times were horrible because the chipset timings are total garbage on that board. I was ruinning aquamark at 3950mhz too with p4gd1(thats on my site to). It's like using fx without a64 tweaker. Not even worth it to bother with pi if going for the top. I could never get over 3950mhz with a p4c/p4p. 18 sec won't happen on any other board but p4c/p4p. You have to have full pat and dram latency parameter 11 this chip seems to run a bit hotter on this p4c. I remounted the container at least 15 times already, with no difference. It could be because there is now alot of surface area missing from the die of the chip around the edges.
18 secs with dothan isn't as easy you fellas think lol...I want it bad and will try very hard believe me. Why dont you guys give it shot ? :)
So much props to japs for making it look so easy...those guys are sick.

As far as my pictures, they are ALL gone becasue something like a month or two ago, my server account expired and I moved every thing over from picture trail that was there to my pc, and all the links went dead. Every single picture I posted here at XS from the very beginning. I emailed or pm'd you and told you this along time ago dude...It would be a pain in the ass to go back and replace every single pic....there are a :banana::banana::banana::banana: load.

Patience AFI....I am working hard at it. I will post it here when I get it, believe me! It will be a big deal for me personally.

[[XC] moddolicous]

Wait, didnt u say that the ES one didnt work as well?? That the retail one was better??

[k|ngp|n]

Wait, didnt u say that the ES one didnt work as well?? That the retail one was better??

Depends on what you call not working as well.
Here is where I am at with the 780 retail chip below.
After working some more with it (retail):
FSB scales nicely with mhz, but the chip is very temp sensitve and really doesnt seem to like below a certain temp. It's the first dothan I ran that is like this. Dothan's NEED every bit of cooling you can give them usually. At 3827mhz with 15x, around 19.3- 19.4sec 1m and good low 4min 8m is still possible :). It doesn't seem to like the high volts like the ES either. Scales and behaves alot like the retail 770m did. The ES's seem to be a diff chip all together.
Good 1m considering cpu clocks ony at 3800mhz:
http://pic8.picturetrail.com/VOL256/1324330/8242801/115906917.jpg

This is new personal record for me for 8m with dothan..score can still get tighter with some more windows tweaking I'd think:
http://pic8.picturetrail.com/VOL256/1324330/8242801/115906923.jpg

[harleybro]

Very nice so far. W/ your determination I am sure you'll pull off the 18's. ;)

[k|ngp|n]

whats wrong with the 8m time?
should be in the 3in time frame for those clocks.
fx57@3.7ghz hits 4m 10s :)

I don't know to tell you the truth. I am not sure how a dothan compares in 8m clock for clock against fx. I dont think it is has the advantage like it does in 1m, but not 100%.
This was my previous best 8m on 770m while back:
http://pic8.picturetrail.com/VOL256/1324330/8242801/110917888.jpg
Pretty sure that previous one was on winxp, but I can't remember if I was running the major tweaks or not and which one(s) if I was.
OPB told me about some tweaks for windows that sig. boost the longer calculation pi benches. Both of these 8ms were basically just run with some services diabled, but no registry, pagefile or maxmem tweaks...which btw I have no idea what the best maxmem for 8m would be, but it would prolly give a nice boost too I would imagine. I was kind of shooting for low 4seconds 8m at a little bit higher clocks.

[G H Z]

Getting closer :)...

Almost 1/2 sec away now:


Excellent run KP :woot:

Is it 670 time yet? ;)

[k|ngp|n]

Excellent run KP :woot:

Is it 670 time yet? ;)

Almost G ;)

[Vegitto]

Is it 670 time now?

[k|ngp|n]

Is it 670 time now?

So many diff. hardware configurations available to have fun with. I wish I had more time for everything :( . Gonna be doing some X dothan single card stuff with a new vid card. Working on getting the bus speeds high and as tight as possible in preperation for 3d on the p4gd1 atm. 3850mhz+ on the big D with 275+ bus speed would be nuts for 01'. I will be testing some new things.

For the record I AM NOT currently selling my dragon cpu unit. There was only 1 fully functional unit ever made, and there are no others that exist. I get alot of alot of pm's and emails about it and I appreciate all of that of course...but for the moment I have NO plans to mass produce it. If I haven't responded to a email or pm..I apologize. I'm not being a :banana::banana::banana::banana:, just been really busy with other commitments recently. I try to answer when I can.

kp

[.sentinel]

Is dragon CPU unit this Solid Container?

[Waus-mod]

mayb the new rev's 780ES for a great fsb @ high mhz :)

[k|ngp|n]

mayb the new rev's 780ES for a great fsb @ high mhz :)

I hope so waus.....The Es's are awesome clocking chips. There will be a flood of them soon on the orbs :).

[Waus-mod]

Hmmm im also gettin me one. i have no expirience with the rev b. but hopefully the rev c im getting doesnt has the fsb problem

[k|ngp|n]

Who wants to see what a 780m and 7800gtx 512 can do?? :)

[Jack]

yeah :D show us pics!

[[XC] moddolicous]

A better question would be: Who doesnt want to see it? Of course we want to see.

[harleybro]

Who wants to see what a 780m and 7800gtx 512 can do?? :)
Ome on post up some screenies! ;)

[k|ngp|n]

Ome on post up some screenies! ;)

Here are some more kick-ass picks from my dragon :P
http://pic8.picturetrail.com/VOL256/1324330/7857526/113860128.jpg

http://pic8.picturetrail.com/VOL256/1324330/7857526/113860121.jpg

Have something else really cool in the works..I will post it up soon.

[[XC] moddolicous]

Cant wait Kingpin. Might we see 51k 3d01 single card?? 4ghz SPI 1m stable dothan??

[metro.cl]

hey kingpin, i decided to give it a go to make a solid container, but i have a couple of questions regardings size (not internally). I would like to know the thickness of the walls (i read it was 3/8) now how wide is the tube (diameter), how thick should be the cooper between the cpu and ln2 dryice? (i mostly bench with dry ice but maybe i could try some ln2 in the future) and how big is the block of cooper before the process. I hope you can help me i already have my first tube but i want a new one

i believe i could get cooper or aluminum at a really low price.

[k|ngp|n]

Hey metro...Good luck with your container man. :toast:
Only measurement I will give out is the upper wall thickness on mine at .186.This would be an excellent starting point for you. You could use that thickness for the base too and prolly get pretty good results....though my base in one area is over 2 inches thick :P

As for width, my pot at it's widest point is less than 2.5 inches and any wider is not necessary IMO.

[k|ngp|n]

Time to start seeing what either of these chips can do on my dragon unit. Pedro's did pretty well that has the same stepping as the first one I tested last night on the left. The # on this chip was cabne 0530 apmw. I tested using the SLI-D as I am not all that thrilled about the nf4 expert so far with my mushkin-bh5:
http://pic8.picturetrail.com/VOL256/1324330/6968932/120542981.jpg
http://pic8.picturetrail.com/VOL256/1324330/8082894/121035434.jpg

After spending an hour or so with it using ln2, It definitely has a cold bug and the mem controller freaks out and I get lots of rounding errors and crashes at bios temps below -33c or so. This chip is very different than my previous week 16 FX. That chip could go forever at -40c in bios.
That being said, I have somewhat figured out a way around this problem and I was able to run the chip at bios/evap temps below -40c/-75c last night and got my best clocks on any FX to date. These runs were done with IHS still on and NO MAXMEM TWEAK :eek: With that tweak, I might have broken 1m AMD WR using 15x lol:
http://pic8.picturetrail.com/VOL256/1324330/8082894/121035424.jpg

I think 3900mhz+ for benches may be possible on this chip with ihs removed, but like I said at one point the mem controller went nuts and I had to raise multi among other things:
http://pic8.picturetrail.com/VOL256/1324330/8082894/121035423.jpg

Ram or mobo are not the problem, it is definitely the chip at certain temps. On Air it goes pretty tight:
http://pic8.picturetrail.com/VOL256/1324330/8082894/121035428.jpg
http://pic8.picturetrail.com/VOL256/1324330/8082894/121035426.jpg

The smoking gun :P
http://pic8.picturetrail.com/VOL256/1324330/8082894/121035422.jpg
Going to grab some more ln2 in abit and see if can take some sp 1m or 8m AMD records with this apmw chip first, then I will try the 0528 gpmw and see how that one does....

[metro.cl]

Hey metro...Good luck with your container man. :toast:
Only measurement I will give out is the upper wall thickness on mine at .186.This would be an excellent starting point for you. You could use that thickness for the base too and prolly get pretty good results....though my base in one area is over 2 inches thick :P

As for width, my pot at it's widest point is less than 2.5 inches and any wider is not necessary IMO.


thanks mate that ware all the measures i needed, im definetly goin to make some containers (i'll begin with aluminum first i believe to test some internal designs then i might go for cupper ).:toast:

update: i just called to ask prices and a 1mt 2 1/2 inch piece of aluminun is 30usd and the same in cupper is 130usd, i'm goin for an aluminum pot :)
another quick question what is the material of the hold down?? how much preasure a cpu with ihs can take?

regards

[gocchin]

Good stuff Vince... I'm watching your 0528 GPMW results closely as I have the same stepping ;)

[k|ngp|n]

thanks mate that ware all the measures i needed, im definetly goin to make some containers (i'll begin with aluminum first i believe to test some internal designs then i might go for cupper ).:toast:

update: i just called to ask prices and a 1mt 2 1/2 inch piece of aluminun is 30usd and the same in cupper is 130usd, i'm goin for an aluminum pot :)
another quick question what is the material of the hold down?? how much preasure a cpu with ihs can take?

regards

Delrin is the best material for holdowns IMO. It's very dense and rigid and can be machined easily. It provides some nice spring on the container without over flexing. You will be very happy with it :)

I got sicker than a dog two nights ago, so it might be a day or two here until I feel better to try that other chip out Tario. I'll let you know soon how it does.

[Cpt Twitchy]

Why no insulation on top half of tube?

[metro.cl]

Delrin is the best material for holdowns IMO. It's very dense and rigid and can be machined easily. It provides some nice spring on the container without over flexing. You will be very happy with it :)

I got sicker than a dog two nights ago, so it might be a day or two here until I feel better to try that other chip out Tario. I'll let you know soon how it does.


Thanks mate i called a store and a solid piece of aluminum with this dimensions:

1 meter long, 2 1/" inches diameter costs: 35 usd.

samething on copper: 120usd

so i might do several internal designs and test them in aluminum, do you have any ideas i might send you one if it works good (you pay the shipping :)) to beta test them, i believe here should be way cheaper to make.

[gocchin]

I got sicker than a dog two nights ago, so it might be a day or two here until I feel better to try that other chip out Tario. I'll let you know soon how it does.


Ah sorry to hear you're not feeling well Vince... it's definitely gotten a lot colder these last few days hasn't it? Get well soon, the cpu testing can always wait till you're 100% :).

[harleybro]

Hmmm a lesson to be learned. Play with LN2 too much and you too can get the cold bug. Sry to hear V. On the other hand for just starting with the first chip it definately seems like you're off to a good start. Wish you a speedy recovery to get bk and start takin more records. :toast:

[G H Z]

I just got over a cold :p

Dang KP you really cranked on that FX. Those are some rippin PI runs, so close to the 21 sec zone. Has anyone done that on AMD yet?

Well I hope your feeling like yourself soon so we can see some more crazyness :)

[ilkkahy]

Thanks mate i called a store and a solid piece of aluminum with this dimensions:
1 meter long, 2 1/" inches diameter costs: 35 usd.
samething on copper: 120usd

1 meter long 64mm solid copper piece only 120$? Thats so cheap man.. here as new, measily 1m long 64mm, 2mm thick PIPE costs 50-60e. Dont know what solid piece would cost but too expensive anyway (+cant be found anywhere).

[k|ngp|n]

Why no insulation on top half of tube?

Thats how I run it when I install drivers and programs or change settings. No need for insulation on the top half of the tube. I just hit it with little splashes every once in awhile and I can maintain 20c container temps :)
I just thought it looked cool when the smoke creeps out over the shiny copper. Top insulation goes on when I take the container below 10c.

And thanx for all the get wells guys :clap:
I am still sick as hell though and it seems to be getting worse. I spent all day yesterday in bed :( and have been up since 2:am.

Metro I would be more than happy to test out anything you make and give you proper feedback on it. I would pay the shipping. I have ideas for new pot designs. I was to start on a new special one this week, but my body had other ideas.

[Waus-mod]

Get well soon kp... when youre better we will see a 3900+ bench i bet!

[T_M]

Kingpin, how come you dont appear to have the floppy or molex power cables connected to the mobo? is that also just while you are messing around at "high" temps?

[k|ngp|n]

Kingpin, how come you dont appear to have the floppy or molex power cables connected to the mobo? is that also just while you are messing around at "high" temps?

I don't plug the floppy mobo power conector unless I am benching SLI usually. The 12v molex plug sometimes I plug in when running 2d, but not really necessary when setting up or installing stuff before bench sessions. The DFi's seems to run fine at stock speeds without them plugged in.

[s7e9h3n]

...That being said, I have somewhat figured out a way around this problem and I was able to run the chip at bios/evap temps below -40c/-75c last night and got my best clocks on any FX to date...
What did you do to get around that prob? Also, the cold-bug seems to scale with temps. By this I mean higher htt's are harder @ lower temps. So around what FSB were you running your cpu @ when it bugged at -40C? My 0512 FX57 will freak out @ 300HTT and ~-42C Core temp (don't ask me why I was trying to run 300HTT with a FX57 - I just was :p: )...Hope you feel better soon :toast:

[k|ngp|n]

What did you do to get around that prob? Also, the cold-bug seems to scale with temps. By this I mean higher htt's are harder @ lower temps. So around what FSB were you running your cpu @ when it bugged at -40C? My 0512 FX57 will freak out @ 300HTT and ~-42C Core temp (don't ask me why I was trying to run 300HTT with a FX57 - I just was :p: )...Hope you feel better soon :toast:


Starting to feel a little better today...Damn I was messed up :/.
Got my second 512gtx modded up this morning and I will start off with some single stage temps (-40c) on the cpu and the cards on water for 3d sli with the 0530 apmw.

As far as the cold bug, I dont even think this chip can run straight DI/acetone without problems like my 0516 could. It freaks on me below -33c evap temps with IHS on. Some things I did to get the chip working at temps below -80c evap/-43c bios were to give it 1.8v, ricky tweak, and mostly I couldn't let the thing sit there idling in windows without getting massive errors and eventual shutdown. If I keep the chip "working" like sham says, I can run it pretty cold....even colder than the 0516 which couldnt take much more -43bios/-80c evap no matter what.
I have only spent a little over an hour with it so far...more time with it will yield some good things I suspect ;)

[s7e9h3n]

Starting to feel a little better today...Damn I was messed up :/.
Got my second 512gtx modded up this morning and I will start off with some single stage temps (-40c) on the cpu and the cards on water for 3d sli with the 0530 apmw.

As far as the cold bug, I dont even think this chip can run straight DI/acetone without problems like my 0516 could. It freaks on me below -33c evap temps with IHS on. Some things I did to get the chip working at temps below -80c evap/-43c bios were to give it 1.8v, ricky tweak, and mostly I couldn't let the thing sit there idling in windows without getting massive errors and eventual shutdown. If I keep the chip "working" like sham says, I can run it pretty cold....even colder than the 0516 which couldnt take much more -43bios/-80c evap no matter what.
I have only spent a little over an hour with it so far...more time with it will yield some good things I suspect ;)
Ahhh...OK...best solution to the cold bug I found - use a 0512MPMW ;)
http://img266.imageshack.us/img266/3576/467ir.jpg

[Dumo]

KP, which one better 0528 or 0530?
You're right with 0516 and Ricky;s tweak I can run a little better @-89 evap. max..... Colder than that and it will stop. It kinda sucks to have only "less than a minute" window to push the cpu after it boot to windows:(

http://img530.imageshack.us/img530/6154/screenshot0105ew.jpg

[[XC] moddolicous]

Kingpin, easy way to keep the cpu busy while idle in windows would be make it crunch d2ol for xtremesystems. SoF does it:
http://www.xtremesystems.org/forums/showthread.php?t=82170
Cant wait for the 3d benching!

[afireinside]

Just run SP2004. D2OL takes time to load plus you need something to keep it at idle like XWD. SP2004 takes less space and is better. His D2OL cands would never get uploaded anyway.

[k|ngp|n]

Just run SP2004. D2OL takes time to load plus you need something to keep it at idle like XWD. SP2004 takes less space and is better. His D2OL cands would never get uploaded anyway.

Thanx for the tips AFI...

Her are some pics of the container I made/use for the middle card on the tightly spaced DFI. It was constructed from dissambled dangerden and swiftech water blocks and some copper thin wallled brazed pipe. It worked very well with gtx256 and so far so good with the 512:

It only works with ln2, as there isnt barely any room for enough DI to cool the card. I use foam gaskets around the core area front and back with my cards:
http://pic8.picturetrail.com/VOL256/1324330/8082894/121808619.jpg

http://pic8.picturetrail.com/VOL256/1324330/8082894/121808615.jpg

http://pic8.picturetrail.com/VOL256/1324330/8082894/121808618.jpg

[Waus-mod]

pretty work kingpin :) gj way to go :D

[harleybro]

Nice V :toast: Still like the other idea better. ;) BTW looks like I might be giving you a ring in the next few days w/ a tube idea to try. :)

[k|ngp|n]

Nice V :toast: Still like the other idea better. ;) BTW looks like I might be giving you a ring in the next few days w/ a tube idea to try. :)

"other " idea is still happening J
I talked to my boy with the cnc this morning about a few things. I will post it here when first one is done. I hate multiple threads on same topic.
Should have some new work with the SLI setup shortly too, no time for anything lately....

Btw-- I managed 780/2050 with that silly little container yesterday on one MSI 7800gtx. The xfx card goes to 820/2100+ at same temps :(

[gocchin]

Those are still pretty awesome clocks... don't you hate it though when the two cards are so unequal? My case is even worse, both XFX, one does 746/1980 and the other 676/1905 individually on water....

Getting ready for you to smash some 05 and 03 records!!! :)

[k|ngp|n]

Those are still pretty awesome clocks... don't you hate it though when the two cards are so unequal? My case is even worse, both XFX, one does 746/1980 and the other 676/1905 individually on water....

Getting ready for you to smash some 05 and 03 records!!! :)

Damn that does suck tario..the cores are way different too :(
Are you going to try and get a better card to a match the good xfx? These cards are nowhere to be found. I might be able to get close to 800 out of the core on the msi card for 03', but the mem isn't going over 2060 or so for sure. It was artifacting pretty bad and more volts don't help. My xfx card is insane. I think it's capable of something like 830-2130+ for 3d with the right touch. I definitely need another card to match it.

[gocchin]

Damn that does suck tario..the cores are way different too :(
Are you going to try and get a better card to a match the good xfx? These cards are nowhere to be found. I might be able to get close to 800 out of the core on the msi card for 03', but the mem isn't going over 2060 or so for sure. It was artifacting pretty bad and more volts don't help. My xfx card is insane. I think it's capable of something like 830-2130+ for 3d with the right touch. I definitely need another card to match it.

Yeah Vince, it seems like if you luck out and get an awesome XFX card then it's usually a killer card. I think the one strong XFX I have will rock with sub zero cooling... as a matter of a fact I'm running both cards only with vgpu mod, no mem mods at all as the memory doesn't seem to respond well at all to any increase.... 830/2130 would be a dream.... I'm stuck with this pair for now cause as you said, they are just nowhere to be found and they're not exactly cheap ;)

[k|ngp|n]

This apmw chip might be the one to get the job done, though I've yet to test the gpmw :)
Ran some mark 05 and 03 single card (def gtx clocks) this morning with -82c on my dragon unit and IHS still on and got this using 15x:


http://pic8.picturetrail.com/VOL256/1324330/8082894/122482559.jpg

http://pic8.picturetrail.com/VOL256/1324330/8082894/122482558.jpg

14x was good upt to 3750mhz, then I had to raise multi due to mem controller taking a :banana::banana::banana::banana:. Pretty sure there is still more left in the chip, especially with some colder base temps. I had the vcore at 1.525x 123% and my cpu unit is still killin' it :P

[k|ngp|n]

Say hello to my little friend..
No more ln2 supply issues for me. Dropped off 230 litres of Xtreme glory. Now I truly have an unlimited supply of ln2. Will definitely come in handy soon.
http://pic8.picturetrail.com/VOL256/1324330/8082894/123099784.jpg

[Willis]

Say hello to my little friend..
No more ln2 supply issues for me. Dropped off 230 litres of Xtreme glory. Now I truly have an unlimited supply of ln2. Will definitely come in handy soon.
http://pic8.picturetrail.com/VOL256/1324330/8082894/123099784.jpg
LOL
:D

[harleybro]

Beautiful thing isn't it? ;) How tall is your container it looks much shorter then mine but mine is supposed to be 180L? Is it high pressure or low. Mines low at I believe 22psi and I would guess about 6ft tall.
http://img224.imageshack.us/img224/8731/picture1244hv.th.jpg (http://img224.imageshack.us/my.php?image=picture1244hv.jpg)

[s7e9h3n]

This apmw chip might be the one to get the job done, though I've yet to test the gpmw :)
Ran some mark 05 and 03 single card (def gtx clocks) this morning with -82c on my dragon unit and IHS still on and got this using 15x:


http://pic8.picturetrail.com/VOL256/1324330/8082894/122482559.jpg

http://pic8.picturetrail.com/VOL256/1324330/8082894/122482558.jpg

14x was good upt to 3750mhz, then I had to raise multi due to mem controller taking a :banana::banana::banana::banana:. Pretty sure there is still more left in the chip, especially with some colder base temps. I had the vcore at 1.525x 123% and my cpu unit is still killin' it :P
Wow, that's a nice cpu you've got there. I only wish my chip would take the voltage. It refuses to boot into windows @ 1.74+ Vcore even though evap temps are ~-72C :rolleyes: BTW, I'm not sure which bios your using with which board but stay away from anything higher that +113% for the Vcore. You'll significantly increase the chances of killing your cpu ;) But then again, 1.88Vcore isn't an option in the bios :p:

[k|ngp|n]

Beautiful thing isn't it? ;) How tall is your container it looks much shorter then mine but mine is supposed to be 180L? Is it high pressure or low. Mines low at I believe 22psi and I would guess about 6ft tall.
http://img224.imageshack.us/img224/8731/picture1244hv.th.jpg (http://img224.imageshack.us/my.php?image=picture1244hv.jpg)

Very :)
It shorter than yours but alot fatter. I was going to get 180litre, but the 230l with the welded on cart is easier for me to wheel around and a little more capacity. It weighs around 1100 pounds full I think.

s7e9h3n: thanx, It seem like a great one so far. The really good ones to me are the chips that can take the volts, because they almost always scale the best. So far all runs done on SLI-DR...can't remebr what bios. Should be benching the expert/SLI on ln2 soon.

[afireinside]

:slobber: You guys have no idea how envious I am of those massive ln2 dewars you have. The worlds best cooling whenever you're in the mood to bench!

[Cpt Twitchy]

Beautiful thing isn't it? ;) How tall is your container it looks much shorter then mine but mine is supposed to be 180L? Is it high pressure or low. Mines low at I believe 22psi and I would guess about 6ft tall.
http://img224.imageshack.us/img224/8731/picture1244hv.th.jpg (http://img224.imageshack.us/my.php?image=picture1244hv.jpg)
I saw 3 of these yesterday at My uncles work. I don't even know why they had them as its a produce company. I saw them in the food processing department. Anyone got an idea as to why they would have these. My uncle doesn't know why as he pick produce order and my bro doesnt know either(he works there aswell).

Tim

P.S. Oh and today my dermatologist sprayed some LN2 on my skintags. It was really cool.

[k|ngp|n]

:slobber: You guys have no idea how envious I am of those massive ln2 dewars you have. The worlds best cooling whenever you're in the mood to bench!

I was in the garage filling my 30l dewar up last night after 2am sometime :)
I left the door open for ventilation and the smoke was rolling out onto the driveway. My neighbors are prolly like wtf is with that guy...lol

[Dumo]

I was in the garage filling my 30l dewar up last night after 2am sometime :)
I left the door open for ventilation and the smoke was rolling out onto the driveway. My neighbors are prolly like wtf is with that guy...lolMad scientist @ work:toast:

[Willis]

I was in the garage filling my 30l dewar up last night after 2am sometime :)
I left the door open for ventilation and the smoke was rolling out onto the driveway. My neighbors are prolly like wtf is with that guy...lol
:rofl:

you prolly bomb them with dryice ass wel :D

[s7e9h3n]

:rofl:

you prolly bomb them with dryice ass wel :D
No....that's what comes out of the bathroom...lol

[s7e9h3n]

:rofl:

you prolly bomb them with dryice ass wel :D
No....that's what comes out of the bathroom...lol

[afireinside]

ROFL! I've thought about bombing neighbors then I remembered he's a cop :D

I remember showing my dad a pic of harleys dewar and saying we need one of these in our driveway... I don't think he liked the idea very much.

[gocchin]

I sometimes dump my leftover DI into the snow outside the house and it starts smoking up... yeah the neighbours said exactly that "What the heck are you doing???" ;)

[Dumo]

I took my phase, mobo etc. on business trip to midwest. I was stayin on one of those extended stay hotel....... You should see the expression on(room service lady) her face :eek: ...the first time the did room service:)

[k|ngp|n]

hehe... we are a crazy breed no? lol

-85c loaded base temps on my container got me this today :)
http://pic8.picturetrail.com/VOL256/1324330/8082894/123434774.jpg
IHS still on too. Card were only run at -25c and still 2140mhz almost!...with lower gpu temps and some more clocks on core, that 18,6 score at the top isn't looking so tough after all. I thihk it can be done on less than 3900mhz. I tested it today up to -98c and it was good running 1.55x123% lol :eek:

[bachus_anonym]

hehe... we are a crazy breed no? lol

-85c loaded base temps on my container got me this today :)
http://pic8.picturetrail.com/VOL256/1324330/8082894/123434774.jpg
IHS still on too. Card were only run at -25c and still 2140mhz almost!...with lower gpu temps and some more core clocks on core, that 18,6 score at the top isn't looking so tough after all. I thihk it can be done on less than 3900mhz. I tested it today up to -98c and it was good running 1.55x123% lol :eek::shocked: :slobber: ehhh...what was I supposed to say...? ...forgot...

[Dumo]

I thihk it can be done on less than 3900mhz. I tested it today up to -98c and it was good running 1.55x123% lol :eek:Nice:toast: ...Still with 0530?

[k|ngp|n]

Nice:toast: ...Still with 0530?

Yep. This chip would probably do well on a cascade too with the just the right heat load dumo.

[k|ngp|n]

Sooooo close tonight.....
Still only around -30c to -40c on the cards. 175 points from 05' WR with only 3860mhz lol:
http://pic8.picturetrail.com/VOL256/1324330/8082894/123601748.jpg

50 points from 03' WR. 44k soon :)
Run was at 740/2144. 800+ on the cores should give around 44.5k
http://pic8.picturetrail.com/VOL256/1324330/8082894/123601745.jpg

[Dumo]

Almost there KP:toast:

[k|ngp|n]

KP>
Set PCIX Freq at 118mhz in the bios ;)
This will give you 50points in 2k5 and 90-100 in 2k3 right off the bat guart. :D

Thanx for the tip...."Lucus" ;) Effect on clockability?

[k|ngp|n]

No effect at all on overclocking. Its just above 118mhz your scores start to go down not up as you saturate the bus. :)

test it on air run 2 2k5's with it at 100mhz and 2 at 118mhz to rule out a lucky run. :)

Thx again. I'll try it in a bit. 123-125pci-x freq was cool on the intel board with 512mb gtx, wasn't sure about the DFI.

[Cpt Twitchy]

You don't even lok like your pushing the cards in 03 :toast: Keep up the nice work.

[k|ngp|n]

Got shamino :D
http://pic8.picturetrail.com/VOL256/1324330/7857526/123716972.jpg
Next up Kinc and Overklokk. Only 60 points more and I got it. Had some minor problems with the cards last session, but I think I got it sorted now. Cards are now bench stable through 05' at 2171+ on the mem. GPU temps were about -45c to -50c or so today. Mother nature is giving me fits again, otherwise I think I would be sitting here at 44k right now. 118 pci-e did help a littlebit, but not great in 05'. I got about 70 points xtra in 03' and maybe 10 in 05' I can take 05' WR with around 3870 on the chip I think. I am bailing on partying later to try for these two WR's tonight. I just cant stop.

[gocchin]

You're the man Vince.... It's going to be a smashing new year for you! All the best!!

[Jort]

That all wishes may come true!!!;)

[G H Z]

I am bailing on partying later to try for these two WR's tonight. I just cant stop.

WTF :confused: ;)

Good luck KP :)

[k|ngp|n]

Happy new year guys! :toast:


WTF :confused: ;)

Good luck KP :)

Well G...I couldnt seem to convince my girlfriend that orb domination was worthy of cancelling our dinner and pary plans..lol.
Will be benching later today ;)

[k|ngp|n]

A little CPU testing with def single card. Getting real close to 3900mhz with 15x for 3d stable. I think this will be good enough for now :O
-92c held for whole run:
http://pic8.picturetrail.com/VOL256/1324330/7857526/123878288.jpg

[k|ngp|n]

Good call. I should grab some up tommorrow....why not. Last tccd I used was gskill long time ago. WHats my best bet for getting some good ones?

[Overklokk]

Dam nice Kingpin :)

Looks like Kinc and i must start upp the rigg again soon :) My FX57 did 01 with singel GTX today @ 3906mhz with -105c load temp on evap . But my w2k instal was bad so only godt about 48473p :(

http://service.futuremark.com/compare?2k1=8847154

Wil reinstall OS and get som better cooling on the GTX in a few days. Think my fx can do about 3930mhz under my 3 stager 3d stable :)

[gocchin]

Nice work V, I see you're almost there in 05 on the HoF.... :)

[k|ngp|n]

Dam nice Kingpin :)

Looks like Kinc and i must start upp the rigg again soon :) My FX57 did 01 with singel GTX today @ 3906mhz with -105c load temp on evap . But my w2k instal was bad so only godt about 48473p :(

http://service.futuremark.com/compare?2k1=8847154

Wil reinstall OS and get som better cooling on the GTX in a few days. Think my fx can do about 3930mhz under my 3 stager 3d stable :)

3930 would be very nice. I prolly won't run 01 with the SLI setup.
Sounds like a battle is coming.... GL with your hardware. You guys are tops on your own...as a team just awsome :) :toast:

[k|ngp|n]

He got me on gt2 by a little too.
2gb scores how much better than 1gb clock for clock?
2gb good clockers on average?

[k|ngp|n]

If you order one of the UCCC kits you should be good for 285-295mhz at low 2.6-2.8v.

Lard. at vr-zone did a UCCC 1024x2 Corsair pc4000 review and he got almsot 300mhz at 3-4-4-8-1t as for how much you gain from 2048 over 1024 at low multi high htt thats unknow as Denny was the only one going that route. :(

Also Im sure a 3400-3500mhz dualcore would push you up a few points. :)

Sucks when your off by under 100 points in a tens of thousands benchmark.

2x512 for now sounds good, thanx for the link. Dual core will be next thing for sure. The numbers look great. If these cards survive the next week, I will run them with a dc chip eventually :) I have some pretty tight solid gpu containers that should be done in the next day or two. I was going to wait till they were finished to bench these cards really hard, thats why the low core clocks. I only needed enough gpu power for the top spots, think I am good now. THis goofy container I made for the middle card has 0 capacity ;) and I dont want to take the volts up to higher than 1.8 on it.

[s7e9h3n]

Nice CPu power. A 13x300 with TCCD combo would give you 100-200points over the 15x 250-260mhz bh-5 combo in 2k5. It would also but less stressful on the CPU becasue of ram voltage/timings so the lower multi should be able to work then.

Denny ran 3800mhz 13x29xmhz and his system power with 2048meg of 3-4-4-8 uccc ram was like 2-3fps faster in gt1 over kinc at 3900mhz+.

Not sure if you have any tccd/uccc sammy there, but something to add to your list of things to try if it comes down to it.
A couple of issues with this though....
1. Can the cpu run 300FSB @ those temps? I have a funny feeling that the cpu's gonna freak out at that speed @ those temps.... :p:

2. If k|ngp|n's using an expert board, there may be some issues running TCCD.

3. That G.Skill TCCD is G.Skill LA's - IIRC, those run @ guaranteed rated timings @ 2t. 1t if you're lucky ;)

[s7e9h3n]

A couple of issues with this though....
1. Can the cpu run 300FSB @ those temps? I have a funny feeling that the cpu's gonna freak out at that speed @ those temps....
A:I say no in the sense that less VDD ram voltage which ties into the on chip ddr controller on the chip thus less stress than 2-2-2-5@3.6v.

2. If k|ngp|n's using an expert board, there may be some issues running TCCD.
A:Again this isn't 100% true. Some Experts have issues, but only with select CPU's which are mostly dualcore chips and select Multi's.

3. That G.Skill TCCD is G.Skill LA's - IIRC, those run @ guaranteed rated timings @ 2t. 1t if you're lucky
Almost all the FX pc3200 can hit 285-295mhz at 2.5-3-3-5-1t 5 users here on XS alone got this high .
Sorry I posted the wrong newegg link before you want this stuff.
http://www.newegg.com/Product/Product.asp?Item=N82E16820231035

1. So you're saying that less VDD = higher FSB achievable in the cold?

2. I have issues with my expert @ 300+ FSB, TCCD, and my 57 :( (Then again I didn't try the x13 mult:p: )

3. OK, that's better...Didn't think that LA's were the new "have-to-get" TCCD's ;)

[Cpt Twitchy]

The best TCCD out there atm is G.Skill FF and The best 2gig kit is Crucial ballistix pc4000. Then again for the best TCCD ever I'm sure there are some individuals around here who have access to the best around ;) i think you know who I'm talking about hehe.

Tim

[s7e9h3n]

A couple of issues with this though....
1. Can the cpu run 300FSB @ those temps? I have a funny feeling that the cpu's gonna freak out at that speed @ those temps....
A:I say no in the sense that less VDD ram voltage which ties into the on chip ddr controller on the chip thus less stress than 2-2-2-5@3.6v.[/B]
Wait a sec, just re-read and did you mean to say VTT not VDD?

[s7e9h3n]

The voltage on the ram that tracks with the on die controller exactly.

Might be VTT I thought It was a VDD voltage. :confused:
VDD = Vdimm, VTT = what you're talking about :p: - 1/2 of VDD....

[[XC] moddolicous]

The best TCCD I've seen were the new Geil One S. SPI32mb stable @ ddr 644 2.5-3-3-7. Seems pretty good to me.

[Cpt Twitchy]

The best TCCD I've seen were the new Geil One S. SPI32mb stable @ ddr 644 2.5-3-3-7. Seems pretty good to me.
Was that OPB? If so I don't think it was Geil:cool:

[vapb400]

Was that OPB? If so I don't think it was Geil:cool:
I think it was AndreYang, I think he beat out OPB a couple days after OPB grabbed the record.

[Cpt Twitchy]

In that case I have no clue hehe. I just remember OPB recently posting a screen of 32Xmhz at 2.5-3-3-6 in one of team italys threads.

[Cpt Twitchy]

Here we go. Check post 13. Not a finished run though oops http://www.xtremesystems.org/forums/showthread.php?t=84316

[bachus_anonym]

He got me on gt2 by a little too.
2gb scores how much better than 1gb clock for clock?
2gb good clockers on average?Don't bother with 2GB kits for 3D benching... 2GB of Samsung UCCC (8-4-4-3.0) or Infineon Rev.C (5-2-3-3.0) will not beat TCCD clock-for-clock. I don't have Crucial Ballistix (8-3-3-3.0) so I can't give you any results with that, but I would not expect it to beat TCCD (6-3-3-2.5-T).

[afireinside]

Recent TCCD has kicked BH5/UTT CH out of the benching memory throne or what? Why not just use a killer 1gb BH5 kit?

[T_M]

Recent TCCD has kicked BH5/UTT CH out of the benching memory throne or what? Why not just use a killer 1gb BH5 kit?

http://www.xtremesystems.org/forums/showpost.php?p=1212092&postcount=237

[bachus_anonym]

Recent TCCD has kicked BH5/UTT CH out of the benching memory throne or what? Why not just use a killer 1gb BH5 kit?I don't want to hijack Vince's thread but as I said in my other thread:
I think that very good clocking 2x512MB BH-5/UTT is still the way to go and better for benching, especially when fighting with severe cold-bug that impairs CPU's ability to go high on HTT :)

[s7e9h3n]

I don't want to hijack Vince's thread but as I said in my other thread:
I think that very good clocking 2x512MB BH-5/UTT is still the way to go and better for benching, especially when fighting with severe cold-bug that impairs CPU's ability to go high on HTT :)

Funny...cuz LucusScott ;) says exactly the opposite:


The cold testing I've done yeilds higher clocks on ram+htt using lower multi's when cold than bh/utt ram that requires higher voltage.

TCCD does run better on the Dfi sli-dr old school board, but there are alot of combos of drive strength and PRE/MAL needed to get it stable close to the same on the expert.

[bachus_anonym]

Steven,

300x13 6-3-3-2.5 will eat 260x15 6-2-2-2.0 (no doubt) but ONLY If you manage to run 300HTT. If particular CPU will do that under very cold temps, then 300MHz TCCD is the way to go. I have not seen anyone running 300MHz on TCCD above 3600MHz so we don't know If cold bug will not kick in @ 3900MHz preventing from going this high on HTT :)

edit: Also, it's true that low multi might get you some better HTT compared to higher one when fighting cold bug, but x13 is not exactly what we would call "low multiplier" anymore ;)

[.sentinel]

CNC normally you can send a cad file to a machine shop and they can make it for you but there is a price.

[Cpt Twitchy]

I don;t know waht a cnc stand for but its an automated mill end/lathe. Some models are millend only some lathe only and then the all in ones. The range new from around $25,000 for the most basic of cnc's. I have seen them as high as 500,000. It was a 6axis 4spindle 12,500rpm beast.

[vapb400]

A cnc machine is seen in the pics in the 1st post. It cuts the material. They are thousands of dollars. You would want to go to a machine shop (look it up in the yellow pages) and give them a drawing of what you want and have them program and make it for you.

it stands for computer numeric control or something like that.

[Cpt Twitchy]

Your welcome and enjoy the learning.

[chilly1]

A cnc machine is seen in the pics in the 1st post. It cuts the material. They are thousands of dollars. You would want to go to a machine shop (look it up in the yellow pages) and give them a drawing of what you want and have them program and make it for you.

it stands for computer numeric control or something like that.
Well that machine is in the hundreds of thousands of dollars.. near to 1/2 million with additional hardware and controls...

[vapb400]

Well that machine is in the hundreds of thousands of dollars.. near to 1/2 million with additional hardware and controls...
WOW:eek2: :eek2: :slobber:

[Waus-mod]

Steven,

300x13 6-3-3-2.5 will eat 260x15 6-2-2-2.0 (no doubt) but ONLY If you manage to run 300HTT. If particular CPU will do that under very cold temps, then 300MHz TCCD is the way to go. I have not seen anyone running 300MHz on TCCD above 3600MHz so we don't know If cold bug will not kick in @ 3900MHz preventing from going this high on HTT :)

edit: Also, it's true that low multi might get you some better HTT compared to higher one when fighting cold bug, but x13 is not exactly what we would call "low multiplier" anymore ;)

High bandwith will only help for 3dmark... Tccd aint gonna kick bh5/utt in hexus pifast or superpi. If so, everyone would bench with tccd, but again the coldbug cant afford high htt. So 2 times why tccd aint great for this kind of benching.

For example 2x 280 2-2-2-5 will kick arse 320 2.5-3-3-7.... Just try it yourself (i did @ pifast and superpi 1m) 320 isnt even enough to beat the 280.... you need more than 325+ to beat it.

[Waus-mod]

Ok i both have 2x256 but will do what i wrote above.. youre right, and i dont know what 2x512 will do .....

[MTP04]

Very nice work Kingpin!!:clap: :clap:

This might be a stupid question but is thermal compound still used on the cpu with DICE/LN2 containers?

[bachus_anonym]

@Waus-mod and LucusScott...

Let's just leave Vince's thread alone here as our discussion does not fit in this section very well... If you want to debate further, I would rather see you guys in my thread - 3DMark2001/03/05 .: vs :. 512MB/1024MB/2048MB - Comparison (http://www.xtremesystems.org/forums/showthread.php?t=84922).

Thanks for understanding :)

[afireinside]

Very nice work Kingpin!!:clap: :clap:

This might be a stupid question but is thermal compound still used on the cpu with DICE/LN2 containers?

Of course, you need some way of transfering the heat ;) Most everyone uses AS ceramique because it works at COLD temps and can be had for cheap in large quantities. I believe kyosen has used AS5 with good results and I've heard a few people say generic white paste actually works best. KP what do you use?

[Dani]

Nice work, Kingpin:toast:

@afireinside:


AS 5:
Thermal Resistance:
<0.0045°C-in2/Watt (0.001 inch layer)

Thermal Conductance:
>350,000W/m2 °C (0.001 inch layer)

Average Particle Size:
<0.49 microns <0.000020 inch

Extended Temperature Limits:
Peak: –50°C to >180°C Long-Term: –50°C to 130°C

AS Ceramique
Thermal Resistance:
<0.007°C-in2/Watt (0.001 inch layer)

Thermal Conductance:
>200,000W/m2.°C (0.001 inch layer)

Average Particle Size:
<0.38 microns <0.000015 inch
( 67 particles lined up in a row equal 1/1000th of an inch. )

Temperature limits:
Peak: –150°C to >180°C Long-Term: –150°C to 125°C


Can AS 5 be good with xtreme low temps, dont think so

[Waus-mod]

AS5 will get "hard" and the thermal conductivity will be bad! Use silicon or Ceramic.

[MTP04]

Thanks for the replys guys:)

Good info there Dani, I thought that might be the case with AS5.

[s7e9h3n]

Of course, you need some way of transfering the heat ;) Most everyone uses AS ceramique because it works at COLD temps and can be had for cheap in large quantities. I believe kyosen has used AS5 with good results and I've heard a few people say generic white paste actually works best. KP what do you use?
He did - until I suggested he use AS ceramique - now check his new OC's ;)

[[XC] moddolicous]

By Dani's chart, AS5 should be OK for something like Mach2 or so. 150 for long term use, thats about what some mach2's do. Maybe not a modded mach2.

[k|ngp|n]

Of course, you need some way of transfering the heat ;) Most everyone uses AS ceramique because it works at COLD temps and can be had for cheap in large quantities. I believe kyosen has used AS5 with good results and I've heard a few people say generic white paste actually works best. KP what do you use?

I used to used the cheapy stuff...but now I use ceramique faithfully. Once I started to bench ln2 on a reg basis, I realized that the thinner paste gets thin and watery and tends breaks down after regular freezing. With the ceramique, I dont have to break down the conainer/insulation to remount as much due to temps from the paste getting bad. It stays thick longer.