knopflerbruce's K8 fun corner!

Oh well, I always wanted to wait until I had like a gazillion insane clocks to show off before making a thread, but because NO bench session can take place without problems that probably won't happen:p: So... I'll just post what I got this weekend, and add more stuff later. I guess that's the best way to do it:D But first:

Special thanks to:

Michael and Jasmin for being my "CPU slaves" in Germany.

Mr. X for being my "CPU slave" in the States. He should also be given credit for his taste in clothes as he shipped me a VERY nice t-shirt with an AMD logo on it:D (I'll see if I can post a pic of that one later)

Imsochobo for fixing dryice, a location and lots of other stuff I don't remember right now.

The benching was AM2 only this time, no time for 939 action - will do that later;) Since AOD didn't work at all with my LP UT M2R I had to use an ASUS Crosshair and a Gigabyte GA-MA790FX-DQ6. However, the Gigabyte didn't like vcore above 1.6v and can't feed the memory sticks with more than 2.45v or so, so I had to do it old school style with the NF590 board:p: WHich wasn't so easy, as I kept getting bluescreens because of memory problems. Reason: new BIOS wouldn't give more than 1.9v to the sticks, no matter how high you set the voltage. After spending even more time figuring out how to use the /f command while flashing, it was FINALLY possible to run some benchmarks. Here are the scores:

LE-1600 @ 4022MHz http://valid.canardpc.com/show_oc.php?id=492744 WR
LE-1640 @ 4017MHz http://valid.canardpc.com/show_oc.php?id=492745 WR
LE-1620 @ 3990MHz http://valid.canardpc.com/show_oc.php?id=492734 2nd
X2 5200+ Windsor @ 3906MHz http://valid.canardpc.com/show_oc.php?id=492742 WR
X2 5000+ Windsor @ 3763MHz http://valid.canardpc.com/show_oc.php?id=492746 WR
BE2400 @ 3571MHz http://valid.canardpc.com/show_oc.php?id=492157 WR
3200+ Orleans @ 3277MHz http://valid.canardpc.com/show_oc.php?id=492732 WR

All of these were on water, except LE1640 (dice), but with snow/ice in the bucket I used as a res:p: No sub-zero temps, but the scores were pretty good anyway. Not so happy with that LE-1620, but I think I can lower the temps just a bit more, as I could run pretty high on 11 multi. Just 5-10 degrees would be enough to kick it above the limit;)

Some other tests I did:

Pifast:

LE-1600, 31.61s http://hwbot.org/result.do?resultId=815532 WR
X2 5200+ Windsor, 32.47s http://hwbot.org/result.do?resultId=815543 WR

Superpi 1m:

LE-1600, 21.687s http://hwbot.org/result.do?resultId=783943 WR
X2 5200+ Windsor, 22.172s http://hwbot.org/result.do?resultId=815542 WR

Superpi 32m:

X2 5200+ Windsor, 19m 38.078s http://hwbot.org/result.do?resultId=815545 WR

I ran out of time (had quite a few chips to test to see if they had potential), so didn't have time for more 32m. However, I think the 790FX boards are a bit faster than this NF590 so perhaps I'll get better times if I can make the DQ6 work properly;) I did more pifast/superpi than this, but these are the best. Even some wprime32m, but I need vista to get some nice times there. I need like 350mhz more in xp lol.

PS: there are like a million drive strengths in the BIOS of this board, but I have really no clue about what I should set them to:p: Anyone knows what works best for d9gmh/d9gkx? These values go from like 2.00x down to 1.00x (something like that).

reserved

Good job, mate:clap: :clap:

Looking forward to ours dice-weekend:D

wellm, we are having serious issues with motherboard and alike, me and knopfler.

This is what we actually use most time on.
Getting pll's and applications that support them that doesnt use too much cpu.
msg me or knopfler for apps that is good like clockgen alike that are supported on 790fx/gx sb600 sb750 and that series motherboards.

voltage settings, on BOTH our motherboards, and in fact the pwm on the gigabyte motherboard smells strange, so im saving that board mostly for the oc weekends, like this we had.
I also got a dead gigabyte ma790fx-dq6 which we used abit in this session, with dead pwm, while the dfi handles those temps and voltages just fine :)

Bios for the crosshair was kind of flankey.

So thats what we use alot of time on, if we do our cool cooling methods to practice, diagnose, gain experience with the motherboards prior to dry ice, we would save alot of dice and get better runs in the future :D

Pictures are getting uploaded to a web server im in charge of, and more info about our rather strange alternative cooling methods.
And of course knopfler with his amd t-shirt... *droool*

Congratz on WRs!!

I like your approach of more is better :p:


I'm using DQ6 in my PC and I have no problems at all. Granted I've went only up to 1.8V on my old X2 6000+, but still. HTT wise I was able to push around 440-460HTT, don't remember precisely but it is in one of threads in AMD section.

Keep up good work guys!

Quote:

---

Originally Posted by Lightman

Congratz on WRs!!

I like your approach of more is better :p:


I'm using DQ6 in my PC and I have no problems at all. Granted I've went only up to 1.8V on my old X2 6000+, but still. HTT wise I was able to push around 440-460HTT, don't remember precisely but it is in one of threads in AMD section.

Keep up good work guys!

---

It doesn't want to boot with anything close to that. Maybe it was the CPU, but that would be strange... I've had "a few" and only an FX60 I tried had that kind of behaviour at pretty low voltage.

You're right about "more is better", I have alot of funt with these $25 LE chips. Why buy an i7 when you can have the same fun for a FRACTION of the price?:p: Hopefully there will be more dice next time, these F3's need subzero to really shine... even if they do pretty well at like 5c too;)

very nice bro, keep pushing :D

Quote:

---

Originally Posted by Lightman

Congratz on WRs!!

I like your approach of more is better :p:


I'm using DQ6 in my PC and I have no problems at all. Granted I've went only up to 1.8V on my old X2 6000+, but still. HTT wise I was able to push around 440-460HTT, don't remember precisely but it is in one of threads in AMD section.

Keep up good work guys!

---

Thanks :)

the board is fast, but well, rather crinky to work with, it doesnt seem to like the very high voltages.
HT clock can be taken up to 475 on my first board, its on my wall actually, i killed PWM, and that smelled reaaaally bad.
more isnt always better, the 4 phase pwm on the dfi 790FXB m2rsh and lp jr 790gx m2rs is faaar better, and make less heat, and is more stable ;)

there is a lack of settings, even with the advanced settings activated in the bios, and the layout is completly terrible if you ask me though.
but its fast, to some degree, stable and does high htt, almost any board can do 440 nowdays, i would be suprised if we could not.

we tested a bad phenom X4 II 940 and well, im getting a better one!
http://valid.canardpc.com/show_oc.php?id=490976

our cooling methods where snow cooling aswell, yes, you heard me right!

http://joomla.ikt-service:banana::ba...1/IMG_0067.JPG

watercooling, 10 litre water tank, and about a meeter of snow outside, we had a lot of decent cooling at hand, but not too much of the extreme, unfortunatly.
Watercooling was basicly, a iwaki pump, a aquacomputer waterblock and 2 tubes, nothing complex!

This was of course when we ran out of dice :p

Damn, I was testing some BH5 sticks on my CFX3200. On my Venus, no 512mb sticks seemed to like above 3.75v or so, but this board worked fine all the way up to 4v sharp. The only problem was that my FS Epsilon 700w died:p: Lots of spark sounds coming from the power connector on the back of the power supply. I just hope the board didn't die as well - or my x1800 master + 256mb cards I used:( Not to mention the CPU lol.

WOW!

I like snow cooling of CPU :clap:

It's good to hear that new better Phenom is en route to you!


Awaiting more benches :)

Quote:

---

Originally Posted by Lightman

WOW!

I like snow cooling of CPU :clap:

It's good to hear that new better Phenom is en route to you!


Awaiting more benches :)

---

It'll never stop.:D But I'm glad I don't have to bring a PSU, need some RMA now:(

The snow cooling was pretty good, I just wonder what the less coldbugged chips can do at -20;)

Added some photos. That's me, btw;)

great results... some very impressive cpu's there

+ the ram was doing its bit too... good ol'ocz ram :)

haha knut you are great :D kicking asses :clap:

Impressive! :eek:

very very nice results bruce, as always :up:

nice T-shirt too :)

keep pushing.

that is an incredible session and mighty impressive results
some serious chips U have there - shame you didnt have time, coolant to get some other WR with those gems

Grat Results +1 for so much WR's on 1 session. BTW i think you must clean your keyboard xD

You made my day :D
Incredible !!!!

Simply the best, keep waving that AMD flag high :D

Quote:

---

Originally Posted by stunned_guy

You made my day :D
Incredible !!!!

---

Quote:

---

Originally Posted by Oj101

Simply the best, keep waving that AMD flag high :D

---

Thanks m8's:) I'll try to keep your 939 thread alive, but my friend who has the best memory, the best mobo and the best PSU is too lazy to bench atm:p: Hopefully, some juicy AM2 scores will bring him back in the game. My 939 boards always die... his old Venus lives forever;)

Got a cheap FX60 today. CCB2E 0607TPMW, so I doubt it'll fly. But I'll give it a try later, you never know:) If it sucks, I'll just sell it and make like a 150% profit:ROTF:

looking foward to some results bruce. i just nice to see you keep the s939 alive. :up:

What CPU pot you was using?

Quote:

---

Originally Posted by ryba

What CPU pot you was using?

---

Mostly no pot at all:D But when we actually used it, it's the one shown in the photos (perhaps the first picture is the one with the best view of it). No idea what it's called, though. Copper base, aluminum "top". Very nice pot for dryice, at least;)

Fellow be-2400 owner here.

Nice work with the cpu-z WR.:up:

Mine is somewhat of a dud(3381 cpu-z) but I still managed to take both wprime records using air cooling. :cheer2:

Quote:

---

Originally Posted by uncle john

Fellow be-2400 owner here.

Nice work with the cpu-z WR.:up:

Mine is somewhat of a dud(3381 cpu-z) but I still managed to take both wprime records using air cooling. :cheer2:

---

I hate wprime, because I have only XP Pro:p: Vista is SO much more efficient in that bench. But once I have Server 2008, I'll try to kick you off the throne ;) That's a PROMISE!

I hope there are better BE's around, I'm not impressed by my chip either... But I have no idea what brisbane weeks that are good or bad, so maybe I just got a couple of :banana::banana::banana::banana:ty weeks:p:

:)

I hope you like SuperPi1M. :D

I use 770/sb600, xpHomeSP3, and simple 150cfm air. I can't make it any easier for you. Go to work. :comp10:

I'm quite busy these days with work and school, but my new SS is begging me to test some chips all the time:D pi runs etc take too much time, so only suicide shots for now... (when my holiday is here, and I get my sticks from Germany there will be plenty of pi action, too;))

http://valid.canardpc.com/show_oc.php?id=569098 X2 6400+ @ 4269 MHz 4th
http://valid.canardpc.com/show_oc.php?id=565260 LE-1660 @ 4052 MHz WR
http://valid.canardpc.com/show_oc.php?id=568202 Opteron 1218 @ 4042 MHz WR
http://valid.canardpc.com/show_oc.php?id=565265 Opteron 1222 @ 4030 MHz WR
http://valid.canardpc.com/show_oc.php?id=568207 X2 5000+ Windsor @ 3847 MHz WR
http://valid.canardpc.com/show_oc.php?id=568211 FX-62 @ 3836 MHz WR
http://valid.canardpc.com/show_oc.php?id=568209 X2 5400+ Windsor @ 3621 MHz WR
http://valid.canardpc.com/show_oc.php?id=568198 Opteron 1212 @ 3538 MHz 2nd

My 6400+ is garbage, so I can't believe I got it that high:D It stinks on air and water, at least... to my eyes it looks like the others (except the 1212 and the FX-62) are underperforming a bit, justt feel they should go a few MHz higher... Maybe the bad evap mount was the reason, causing uneven temps across the IHS... I'll find out later when I put them on dryice:) That 1212 had a sweeeet coldbug, but sucked on air... maybe I can get a nice score on dryice, I dunno... not feeling too confident about that one.

It's also nice to break Kingpin's FX-62 WR, even if he ran 3D at his old record frequency:clap: (no idea if I could, this unit won't take the heat at 3.6ghz+)

Very impressive stuff. Well done.

Allright... my first REAL LN2 session!:D Had to learn a few things, so it took a little while to get started... what's strange is that this setup seems to love some chips, and hate others. No idea why:shrug: But anyway... this is what I managed with 10L LN2:

3500+ Clawhammer @ 3505MHz 3rd http://valid.canardpc.com/show_oc.php?id=579350
3200+ Newcastle (939) @ 3526MHz WR (by 620MHz:eek:) http://valid.canardpc.com/show_oc.php?id=579347
4000+ Clawhammer @ 3669MHz WR http://valid.canardpc.com/show_oc.php?id=579352
FX55 (it IS an FX55, same batch as the one stephen got years ago - even if CPUZ doesn't agree:rolleyes:) @ 3991MHz 4th (http://www.hwbot.org/result.do?resultId=860268 - was replaced by that "real" FX57 score...)
FX57 @ 4012MHz 14th http://valid.canardpc.com/show_oc.php?id=579354
Opty 156 @ 4030MHz WR http://valid.canardpc.com/show_oc.php?id=579345

I tried a couple of others too, but since I didn't break my personal best (despite better cooling, used dice before), I didn't make any validations.

Not too much time for superpi, pifast or wprime - was pretty hard to control the coldbug, since I had no help, especially since I had no thermal sensor, or thermos (had to pour straight into the pot:ROTF:). But I did a couple of runs anyway, just to see what kind of scores I could get... so I got the FX55 1m WR (with the chip identified as FX57), and the Opteron 156 1m record, which actually is the best score of ANY non-FX socket 939 chip out there:yepp: (5th including the FX series, but a SHORT way up to 3rd place).

Opty 156 superpi1m @ 21.844s WR: http://www.hwbot.org/result.do?resultId=860269
"FX55" superpi1m @ 22.859s WR: http://www.hwbot.org/result.do?resultId=860270

I forgot to put the best score on my memory stick, as it was mislabeled on my bench drive:p: Had to capture it fast to avoid coldbug... if my memory doesn't play games with me, I think it was 22.515s @ 3776MHz.

What an action, very impressive !!!!

You are the men for AMD :D

awesome work bruce :toast:

Got some LN2 today as well (10L). I decided to try some AM2 chips, but my mobo hates me today, so I can barely do suicide shots... Nothing else than auto vcore gives me instant 1.8v, it hates all my 4 memory kits, and it failed to make my Opty 1224 hit 4.3GHz, despite the fact that it did 425x at 1.66v and 5-6C higher than the coldbug:rolleyes: Managed to get ONE validation above 4265, this one: http://valid.canardpc.com/show_oc.php?id=584303

Not too bad, highest CPUZ validated 90nm K8 result (3rd including the 65nm Brisbanes, 4th including the 6400+ WR picture) - but should've gotten #1 on this:shakes: Still, I'll try more. Also got this score on my new 6000+ (125w!):

http://valid.canardpc.com/show_oc.php?id=584305

WR by a mile, and still more in it - I THINK 4250 is possible. But the damn board went crazy before I had a chance to find out.:down:

The last chip was a 4400+ Brisbane, and after struggling for 45 minutes I think I got the record there as well... have to check my benching drive first, when the board from HELL decides to start working again so I can copy the validations.;) 4050-4060 is pretty decent, was at 1.63v and still positive temps:p:

Dunno what to test next... maybe try to grab some HWBoints for that brisbane and 6000+, I'm pretty sure I have a good chance on a few benchmarks, despite using the slow nf590 chipset (compared to 790fx), and XP pro (which sucks on wprime, at least for me).

Very nice Bruce...did you get a thermos? :rofl:

Quote:

---

Originally Posted by miahallen

Very nice Bruce...did you get a thermos? :rofl:

---

Nope:p: I can handle the dewar now, so why bother;)

Testing 939 now... got a couple of HWBoints, too;)

Thats just sick. Awsome stuff!

OMG I hate you, now I have to change my sig :( :( :(

I think my chip is pretty good, it totally owns my Opteron 1222. If it had the same coldbug as the 1224 I would've hit close to 4.3... was like 15c between them:p: I wonder what your new chip, and stunned guy's will do;)

I guess I will test some stuff on water when i get home from work... Feels kinda useless only 12h after I benched with Ln2:rolleyes: But i want my HWBoints:D

Wow man, congrats on so many impressive clocks!

Makes me want to get the dice pot out to see what my Phenom II has in it. :p

GJ bruce. I would be cool if you'd write coldbug temperatures next to every chip benched so people could get a feeling of how CB temperature differs from chip to chip :)

Best Regards,
Niko :up:

Quote:

---

Originally Posted by tiborrr

GJ bruce. I would be cool if you'd write coldbug temperatures next to every chip benched so people could get a feeling of how CB temperature differs from chip to chip :)

Best Regards,
Niko :up:

---

I'll see what I can do;)

CCBCF's bug hard, my LE1620 did 320MHZ HTT at +9c but managed to nail the WR at 13c and 1.79v after one of the hardest battles I've ever had:D

WR LE1620: http://valid.canardpc.com/show_oc.php?id=586988

My LE1600 (also CCBCF) bugged at 13c and 365MHZ when I benched it.

CCB8F's are a tad better than the last one, I guess about 20-25Mhz more at that temperature. Didn't really try too many of those at that high HTT, though.

First results after the summer holiday: http://valid.canardpc.com/show_oc.php?id=654031 FX55 Clawhammer @ 4044MHz WR
http://valid.canardpc.com/show_oc.php?id=652802 3500+ Venice @ 3769MHz WR
http://valid.canardpc.com/show_oc.php?id=651044 3000+ Newcastle (s939 version) @ 3290MHz WR

Nice start, but a few broken mobos as well:(

You are crazy knut ;)
that 3000+ rocks very good.
whats the stepping of that fx-55 ? is this mine cpu ?

why broken motherboards ?

Quote:

---

Originally Posted by stunned_guy

You are crazy knut ;)
that 3000+ rocks very good.
whats the stepping of that fx-55 ? is this mine cpu ?

why broken motherboards ?

---

The same chips as you had, yes.

...and here is your old FX53 in action: http://valid.canardpc.com/show_oc.php?id=654224

It sucked hard on dryice temps, but once it went REALLY cold it wanted to fight it's old record:p: But now I've used enough Ln2 for 130nm, I guess am2 next;)

Broken boards because of WC failure (5L on benching rig while it was running on LN2) and broken socket (technically it still works, but it's too much hassle to remove chips so it's unbenchable.). But it still works for 24/7 (typing on it now:D).

ASUS Crosshair + some Brisbanes next I guess... ;)

New score... http://valid.canardpc.com/show_oc.php?id=654322 Sempron 3200+ Palermo @ 3351MHz WR

...but this one was hard, the temp sensor was all over the place. While loading in windows: 30c, idle: -25c (!).

I wonder why our fx-55 and fx-53 runs better on your board than our board ? all two records are stolen of you :D we ( gianni and me ) thought that will be records for a the future...

Quote:

---

Originally Posted by stunned_guy

I wonder why our fx-55 and fx-53 runs better on your board than our board ? all two records are stolen of you :D we ( gianni and me ) thought that will be records for a the future...

---

I can't explain it... are you sure you got the optimal temps? I know I was close, but because I had no temp sensor I doubt I was at the limit. Also, I seemed to gain 10-15mhz by going from 1.77v to 1.84v on the FX55, so maybe you used too little voltage for that one;) The FX53 was really moody when it got below 145 in ITE, but on a couple of occasions it let me get past 3900.

Could be the board of course. Maybe I'll test the FX53 I got to 376x on on my Expert. I thought that one would kill your record, but it didnt:p: It managed 3560 or something on dryice (on a Venus, not mine), and worked all the way down to 130 in ITE, so 85 degrees lower temps should've given me more than 200MHz extra:p:

Allright... I call it a day now:p: Tried a couple of chips on the Crosshair on air, maybe I'll crack the Opteron 1210 WR tomorrow;) Didn't look to have a too bad CB for these CCB8F chips.

Forgot to mention that the 3000+ NC was an xPBW chip, so they're not THAT bad.:D

Edit: we should've done these OC's while these chips were hot... then we would've been heros :D

awesome score bro :D