Ricky and Pedro Rocha bench session FX55@3867Mhz

[Pedro Rocha]

Finally, with a completely rebuild cascades, we start the bench.

Here is a photo of the bench lab :)

http://pedrorocha.planetaclix.pt/BenchFX55Geral.jpg

17Kg of 99,9% pure ethylene to play with :
http://pedrorocha.planetaclix.pt/EtilenoGarrafa.jpg

Board and CPU ready to start :
http://pedrorocha.planetaclix.pt/DFInaCascata1.jpg

And the first result :
http://pedrorocha.planetaclix.pt/fx3867.JPG


I think it is a new WR :D but I think we can get more than 3.900Mhz with this setup.

Ufortunately the board is quite unstable, maybe because it is an RA00 revision :( so I'm gonna to try to change it for a new revision today..

That stops us to have a perfect run for the our primary target : Numero Uno in 3DMark2001, of course :D

An all tests 3600mhz + run in 3Dmark2001, seems quite feasible with this FX55 and cascades... as alaways Ricky have done a fantastic job : even at 1.9v the CPU never pass -95ºC ..being at near -100ºC most of the time ..

More results in the next days :D

[Pejsen]

w000t :D

[Paxi]

just wow :slobber: !!!!

[mcnbns]

Pedro and Ricky, the dynamic duo to whom no number one spots in any benchmark are out of reach! Huzzah!

[sinsinking]

OMFG.
let's get some bench

[k|ngp|n]

Can't wait for your 3d..... :toast:

[@njo]

Congratulations (Parabéns pessoal)...

As you can see, Portugal its the european country with more sunny hottie days... but not even mother nature can stop us to get cold....

For those you are asking: "what the hell does Aquapc - Quality computers mean???

Check it out here www.aquapc.net, store site, or www.aquapc.com for forum

[eva2000]

god dam awesome :D

[milkycloud]

Congrats on the new record :toast:

I love your cascade :p: But I hate your CRT mon :D

Keep it rolling :toast:

[|RickY|]

i realy wish it is the board... the sistem was very instable, we can get a screen @ 3.9 but it crashes saving, but not dedicate very time to screenshots

[IvanAndreevich]

Yeah that monitor needs upgradin' :) Let's see some 3D @ 3.7?

[SLaY3r07]

Yeah that monitor needs upgradin' :) Let's see some 3D @ 3.7?

LOL, thats for sure. But anyway that FX-55 is very beefy.

[|RickY|]

dont speek bad about my bench monitor :D it needs a kick every 5min to show image but it has shown a good frame rate :D

[SLaY3r07]

dont speek bad about my bench monitor :D it needs a kick every 5min to show image but it has shown a good frame rate :D

Yeah, I know what you mean. I have got one of those downstairs :rolleyes:

[perkam]

omg :slobber: is that bench stable !!!!

Perkam

[skate2snow]

Own did it get downstairs Slayer07?:D....:lol:

And HOW DO YOU WANT TO GET FIRST WITH A NON-CASCADE VIDEOCARD??!??! :D... Not even Vmodded, Come ON!

[skate2snow]

omg :slobber: is that bench stable !!!!

Perkam Perkam, please read the FULL post... If you read it correctly, there is no reason of this question...

[Bennah]

nice work Pedro & Ricky with that FX and the cascade. Seems that the wait has paid off :D

Now torture that x850 ;) Good luck with the 3D Mark 01 benching ;)

Would we see a 3 way battle at the top of 01 orb? OPP, Pedro & Ricky and 'The OC. Godfather Team' ? It would be intresting :D

Looking at the orb, J.Hansen & B.Hermansen score is still up there in 4th and it was done on the 11th Dec 2004 :eek: more props to them for that score :toast:

[[XC] leviathan18]

would you like to change that setup for my sweet 3500 newcastle that crashes after 2.7 xD omg 3.9 stable for 3dmark would be a dream..... bring down the super pi record... amd rox

[|RickY|]

the card on pic is a x850, the x850xt pe is insulated, moded and atached to -100 Gpu cascade, only is mounted when we get a stable system to bench ;)

[Bennah]

oh yea and its time for Pedro to change his avatar :D

[SLaY3r07]

the card on pic is a x850, the x850xt pe is insulated, moded and atached to -100 Gpu cascade, only is mounted when we get a stable system to bench ;)

Can't wait to see that X850 XT PE in action :banana:

[OPPAINTER]

Looking good guys :toast:

Nice CPU clocks. How long before we see some 3D scores?

Nice job.

OPP

[|RickY|]

if problem comes from board, soon

[HARDCORECLOCKER]

:D Conrats guys - AFAIK it is a new WR........, great job !!! :toast:

[musk]

where are the pics? :stick:


oops I see them now.


musk

[bachus_anonym]

I'll check back on when benches show up ;) But you guys are off to a good start and maybe some some 3DM01 @ 3.6-3.7GHz :) So far, nobody has done it as far as I can see.

Good luck! And how about some 3DM03 while you're on it?
:toast: :thumbsup:

[HARDCORECLOCKER]

:D Oh, just noticed You did it with only 1,888 volts - is that the sweetspot of the FX or have You guys room for more with a better board, maybe with 2 volts ? ( let's see the 4GHz...... :p: )

:toast:

[VoRtAn_MaDgE]

I'm not gonna say wow and stuff like that cuzz i was already expecting nrºs like this one's :D

Can't hardly wait for benchs :slobber:

Nice job as always m8s

[maverick89]

very nice score ;)

[DarkManX_BG]

:slobber:
Can't wait for the results :D

[tictac]

Ufortunately the board is quite unstable, maybe because it is an RA00 revision :( so I'm gonna to try to change it for a new revision today..


Maybe you should try benching that with ABIT AN8 or AN8 Fatal1ty

this board use 4 Phase Vcore Regulation Circuit

anyway congrats :worship:

[HARDCORECLOCKER]

Maybe you should try benching that with ABIT AN8 or AN8 Fatal1ty

this board use 4 Phase Vcore Regulation Circuit

anyway congrats :worship:

:confused: But how to get the 276 FSB then with only 2,8 Vdimm ???

Gotcha!!!! :p:

:toast:

[misteroadster]

just Wow ;)
Good job !!!

[Learn]

just awesome guys :toast:
show us more!
:D

[sinsinking]

can't wait for FX-57 to break this record

[HARDCORECLOCKER]

can't wait for FX-57 to break this record

:confused: You own a cascade for that venture...... ? :toast:

[funkflix]

Impressive! :toast:

[SLaY3r07]

How long before we see 4GHZ? :slobber:

[Nukke]

Congratz. I hope you get a good mbo to match with your CPU and your cascade. We wanna see some more WR. :D

Portuguese regards.

Nukke.

[esdee]

:confused: But how to get the 276 FSB then with only 2,8 Vdimm ???

Gotcha!!!! :p:

:toast:

i don't think Vdimm would be a problem ... thats what mods are for


anyway good luck guys! :toast:

[kappa2001]

Vamos para o 1º lugar?
Good job :toast:

[eviLRuLeZ]

impressive :D

[AH^DK]

God damn nice fellas!

[charlie]

amazing...
I can only IMAGINE the 3D POWER.........

But I think you need to update your benching drive... CPUZ 1.27 is OUT! Some of the other WR holders (the Japanese) may not accept a screenie from an older version that's susceptible to manipulation.

C

[HARDCORECLOCKER]

amazing...
I can only IMAGINE the 3D POWER.........

But I think you need to update your benching drive... CPUZ 1.27 is OUT! Some of the other WR holders (the Japanese) may not accept a screenie from an older version that's susceptible to manipulation.

C

:D Ooops, thought 1.27.1 was the latest, will check it... :toast:

[Beat_Slayer]

Congrats guys!

Wayting for bench action! :toast:

[isp]

CPU's got potential then :up:
Hurry up and get a newer revision DFI!! :hehe:

[Patrickclouds]

well done :slobber:

i´m interesting in some 3dmarl 2001 action :toast:

[Playful_Buffalo]

nice job

i await 3d screenies :D

[gloatlizard]

wow great job man

[charlie]

:D Ooops, thought 1.27.1 was the latest, will check it... :toast:

Actually there's a 1.27.2, but the point was that 1.25 can't be used nowadays for a claim at a WR....

[VoRtAn_MaDgE]

Guess that point will not be a problem just download the new one and make a new screenshot

btw the current wr in amd in vr-zone is from memesama and he's using cpu-z version 1.24

http://www.vr-zone.com/ocdb/amd_screenshots/321i.pjpeg

link vr-zone (http://www.vr-zone.com/ocdb/data.php?id=321&intel_or_amd=amd)

[charlie]

Guess that point will not be a problem just download the new one and make a new screenshot

btw the current wr in amd in vr-zone is from memesama and he's using cpu-z version 1.24

http://www.vr-zone.com/ocdb/amd_screenshots/321i.pjpeg

link vr-zone (http://www.vr-zone.com/ocdb/data.php?id=321&intel_or_amd=amd)

yeah, that's because memesama set that old WR several months ago before Franck updated the architecture for more secure/reliable results. I don't think anyne HERE doubts Ricky/Pedro results, but for "mass consumption", it's best to be safe :D

C

[metro.cl]

awsome cpu clocks you both are the best overclockers, you should have been invited to the ati show with macci, fugger and opp.

come one im almost sure you'll brake the 4ghz barrier and the 45k on 3dmark01

hope you dfi can be fixed or changed soon.

p.d. i hate dfi they make great boards but not reliable, thay always have problems.

[Grov]

p.d. i hate dfi they make great boards but not reliable, thay always have problems.

Kind of a silly statement really dude.

My DFI NF3 is awesome best board ive owned, and i owned 2 of the best NF2 boards ever in the NF7-S and Epox 8rda+.

Great CPU clocks man, insane really!!!!!!!!! :p:

[metro.cl]

Kind of a silly statement really dude.

My DFI NF3 is awesome best board ive owned, and i owned 2 of the best NF2 boards ever in the NF7-S and Epox 8rda+.

Great CPU clocks man, insane really!!!!!!!!! :p:



how come its silly, i had a lanparty nf2 was great for oc but always with problems, the new lanparty nf4 also has lots of issues.

[althes]

Great record, just need those benches now.

[Pedro Rocha]

amazing...
I can only IMAGINE the 3D POWER.........

But I think you need to update your benching drive... CPUZ 1.27 is OUT! Some of the other WR holders (the Japanese) may not accept a screenie from an older version that's susceptible to manipulation.

C


Charlie,
CPU-Z 1.27 now on bench HD, thank you for reminde me that :)

I hope next we can save on HD the 3.900mhz screen or more, this CPU really have an amazing power on Ricky's cascade..

BTW : more tests with this DFI revision A00, definitively a lemon :( .. and most probable cause of the instability ..for example even we can run somethimes Lobby at 3.650mhz, superpi dont run after 3.500.. but PIFast run ..thats quite odd..:(

I have a great faith on my Asus X850XT PE - 620-655 with stock cooler and fan at 75%..I hope to attain near 850-700 @ -100ºC - card already fully modded and isolated ..just wainting for the new motherboard :)

:toast:

[Cpt Twitchy]

Awesome clocks :banana: I can't wait to see the your guys' benching results. Keep up the great work and hopefully you'll show us 4Ghz.

[Pedro Rocha]

2 more photos:

The new evap fixation system, quite good ..no condesation at all after 5 hours at more than -100ºC :)

http://pedrorocha.planetaclix.pt/DFI_evapmount.jpg

Here is Ricky modding my X850XT PE :)

http://pedrorocha.planetaclix.pt/RickymodXT850.jpg

[sinsinking]

wow
4Ghz, i'm waiting for it.
but intel 4Ghz, is AMD 4Ghz!!!

[dippyskoodlez]

how come its silly, i had a lanparty nf2 was great for oc but always with problems, the new lanparty nf4 also has lots of issues.

I dont have a single problem with mine.

[sinsinking]

Yeah, Capture some more photo.
U better setup a cam to broadcast for us to see ur process.
LOL

[enzoR]

wow impressive!

[trecca]

extreme clock! :slobber:

[M3taC0rT3X]

:slobber: WoW! :slobber:

Greetz and GJ from Portugal (Porto) to wherever you guys are .. probably Lisbon.

btw: gj aquapc parts :banana4:

[scrible88]

Those are some crazy clocks! Most impressive!

[@njo]

:slobber: WoW! :slobber:

Greetz and GJ from Portugal (Porto) to wherever you guys are .. probably Lisbon.

btw: gj aquapc parts :banana4:

Ricky is from Setubal (VS 0 - 2 SLB, lol)

when will we see this team with an international sponsor...

DFI... send a board to the boys.. you will not regret it... and your stockholders will love it :banana:

[[XC] moddolicous]

Great job guys, those results are insane. We should have a group effort to break 45K. We'll borrow some of OPB's amazing ram, and some other amazing stuff from people.

[STEvil]

F5 with cpu-z to capture screenshot to cpu-z directory.

[skate2snow]

:confused: But how to get the 276 FSB then with only 2,8 Vdimm ???

Gotcha!!!! :p:

:toast: ???? You never REALLY overclock hen? Do you know why we have a high traffic Vmod section;). I think that you have been lucky with the DDR Booster and on board 4V, no?

[LostInSpace278]

Bah, let's see 4 Ghz and -100°C temps :stick:

[isp]

Bah, let's see 4 Ghz and -100°C temps :stick:
Fine, give us an fx57 :slobber:

[WxChaser]

I knew it had been too quite in Portugal for too long - great work Ricky & Pedro, know I'll be following this thread. :toast:

[HARDCORECLOCKER]

???? You never REALLY overclock hen? Do you know why we have a high traffic Vmod section;). I think that you have been lucky with the DDR Booster and on board 4V, no?

:D I know this mate, but TICTAC recommended the ABIT AN8 and AFAIK there is no Vmod available for that board, also the booster won't work I heard so far - thats why people screaming to ABIT "give us more Vdimm"....

:toast:

[skate2snow]

AFAIK (and I know fairly enough about Vmod;)).... There NEEDS to be a chip for Vdimm control and if there is a Vdimm control chip, there is a Vmod... I'll go take a view to that Vmod.

[Agr3sive]

Thats an insane overclock, its enough to tempt any1 to the AMD side.

[Eradicator]

WOW.. amazing stuff here. Keep it up guys. :toast:

[Grov]

So when we gonna have the pleasure of some 3dmark scores? :cool:

[scrible88]

Or PI? Or new screenshots? Or ANYTHING!!!!111oneoneeleven

I'm sorry, I'm anxious! :P

[FUGGER]

Great job guys.

I was that close to the WR...

[jll]

Indeed! Great Job Portuguese Duel! :D

Let us see some marks!

Best wishes

Jorge Lopes

[onemorekornkid]

post some benches!

[enduracell]

AFAIK (and I know fairly enough about Vmod;)).... There NEEDS to be a chip for Vdimm control and if there is a Vdimm control chip, there is a Vmod... I'll go take a view to that Vmod.u make it sound so simple. AN8 imo needs three mods for Vdimm, anyone that had played w/ abit, knows. +it needs BIOS mods for sure->very bad ocability w/ ram like BH and UTT.
Anyway good luck Ricky&Pedro wish u the best :toast:
maybe some BIOS testing will do the trick...

[_King_Pin_]

Bios didn´t solve the problm as I know. The problm really is the mobo (A00).
If they want I´ll gave them my board for some benchies. I just need 30min. to travel between Lisbon and Setubal :D

Força aÃ:banana:! como Vêem tá tudo á espera de novos recordes vossos :D
Good luck For U two!
:banana4: (is this a smiley dedicated to Portugal?? It has our frag color...:D)

[SAE]

Great clocks! Iiimmmmpressive! :D

[zakelwe]

It's been a long wait for you to get your cascade fixed but it seems to have been worth it, almost there now. Main thing is that card and cpu running well on cascades just the "disposable" cheaper motherboard to change out for better one.

I think Opp might be under pressure soon with that cpu speed, video still an unknown though and it's always tough to beat Eric.

Good luck.

Regards

Andy

[@.ru]

Great clock!!! Congratulate!
We want 4Ghz :)

[Apocalipsis]

Just wordless...way to go guys :toast:

[Napoleonic]

when will there be a competition : 4ghz AFX vs 7ghz PEE??? who wins?

[|RickY|]

well bad benching night :S

very problems with condensation...

first the 2 ddr slots dont work.. after 2hours of cleaing it works again...

after some runs of lobby@ 3.64mhz and good damm scores we put the x850xt on action

first tests@ 1.77v /2.11v an amazing dragotic run @ 830/730

after this test can't even run dragotic @ 600 core with 2v i thing it is condensation, card is cleaning now... maybe more soon :S :S

http://www.teampuss.com/forums/files/condensa__o_168.jpg

[jrw]

Keep on fighting! :D

[saaya]

whoohoo, go for the big 4 guys, go for the big 4! :toast:

[tictac]

ic.... TwinMOS memory on your rig :D

[G H Z]

I hope it's condensation:)

You're definately making things interesting, good luck :toast:

[$0m#0n#]

when will there be a competition : 4ghz AFX vs 7ghz PEE??? who wins?
FX! :D

[Paxi]

Good luck! can´t wait so see some results..

[Bennah]

ic.... TwinMOS memory on your rig :D

trust tictac to notice that :D

they UTT or the new BH-5 btw or maybe some old skool BH-5 :D

[madgamer]

Sick overclock on the cpu, but the overclock on that GPU is really out there as well. Thats some serious cold on a graphics card. I cant wait to see some benches with this rig.

[detonator]

well bad benching night :S

very problems with condensation...

first the 2 ddr slots dont work.. after 2hours of cleaing it works again...

after some runs of lobby@ 3.64mhz and good damm scores we put the x850xt on action

first tests@ 1.77v /2.11v an amazing dragotic run @ 830/730

after this test can't even run dragotic @ 600 core with 2v i thing it is condensation, card is cleaning now... maybe more soon :S :S

http://www.teampuss.com/forums/files/condensa__o_168.jpg
gona cry... we having esacly the same problem under cascade , we made every kind of insulation with saran wrap without ,even a box fealed with co2 to kill the condention 5 houars in row of test 0 Condension, but no way to get a rid of this problem.. I hope you will get a rid of this ,so you can tellme wath was the poblem...
goodluck!

[[XC] moddolicous]

Something to try (not sure if its possible) is create a vacuum of some sort to suck the water vapor out?? Is that even possible.

[detonator]

Something to try (not sure if its possible) is create a vacuum of some sort to suck the water vapor out?? Is that even possible.I did that, but dident worked......fealed the sealed box with co2....

[gclg2000]

I know this is a crazy idea....but what about using spray foam insulation on the video card. Not so much directly onto it, but somehow to make a thin "chamber" to put the card into and the walls of the chamber would be spray foam insulation covered in some 1/2" neoprene.... Try to minimize the area of air around the card period.

Create a cardboard "sliver" or something to put the card into that you could press down into some neoprene around the card socket and on the backside of the mobo socket were the Card slot is. It's hard to describe what i'm saying, maybe i could draw a MS paint picture to help understand better or something??

Would be balls to the wall if you covered the card in insulation w/ the evap attached though.

[don_vercetti]

could you not simply surround the whole rig in either a vacuum (although then you'd have to cool stuff like mosfets and everything somehow) or some sort of gas that couldn't vaporise.

Failing that, what's the possibility on immersing it in non-conductive fluids? that would sort out any condensation, quick sharp. Thinking about it, getting that to high (low?) sub-zero temperatures wouldn't be easy, unless there's a non-conductive anti-freeze (i'm not sure about that).

[G H Z]

I did that, but dident worked......fealed the sealed box with co2....

So you had no condensation problems after this but the card simply did not clock high anymore?

Weird....

I think that sometimes just kicking a brand new piece of hardware with voltage/cooling/extreme clocks without giving it a chance to run a little is not sooo good.

[SLaY3r07]

I wonder how good of an overclock you would get w/ the Sapphire Xpress 200 board...

[detonator]

So you had no condensation problems after this but the card simply did not clock high anymore?

Weird....

I think that sometimes just kicking a brand new piece of hardware with voltage/cooling/extreme clocks without giving it a chance to run a little is not sooo good.
esacly, after 5 houars, no condention at all ,but we cudent OC the card, this card can do at stock air cooling with 1.7V on the core and 2.20 on the ram 660/670 passing without any problem the nature.

[Beat_Slayer]

Maybe it's another thing, like a mosfet that is geting more cold than it can handle! :stick:

[|RickY|]

detonator, but your card have done a good oc?

on first 5min ofd benching the card do 830/730 with only 1.77v on gpu after a crash never more pass 650mhz core +- mems still ok

tomorrow i'll test it, it is on clean job :D

[charlie]

dunk it in oil :D
I've already PROVED that FX55 are good til -135C.........

C

[detonator]

detonator, but your card have done a good oc?

on first 5min ofd benching the card do 830/730 with only 1.77v on gpu after a crash never more pass 650mhz core +- mems still ok

tomorrow i'll test it, it is on clean job :D
well basicly, we tested first with the mod off (to see if the contact with the card was perfect) , first run on nature (mod OFF) at 715/650 at -95° then we turned On the mod, but we culd barely bench at 680/600, with the mod on was worst.....and this more then 1 time, like you say, we though about the condention, but every time ,when we opened the card from the wrap or neopreme was dry,we getting crazy.... good luck!

[Pedro Rocha]

Thank you all

Have to use my laptop for posting, almost all my HW on the cascade

Good news :banana:
New motherboard A01 is OK
We got a fantastic Lobbys and CH runs at 3.650 with X850 on stock (in fact 3 new WR on that tests) :)
Memory can handle 277 2-2-2 or more, I saw a new PIFast WR ..less than 34sec :)
But our primary target was 3Dmark2001 ..of course ;)
(when I talked about PIFast, Ricky just tell "I don't give a shi*** about that thing, did you see that fu**** CH run ..f****")

Bad news :confused:
X850 behaviour on -100ºC
After seen detonator post, I start to suspect that the X850 don't like extreme low temperatures.
In the ORB I see 2 good X850 runs (MMouse with gpu swiftech water cooled with chilled water at -30c / mikead_99, 1.7Vgpu, Mach 1) but no good runs on -100ºC..
Maybe it is not condesantion related, but a temperature related problem specfic to X850..
Macci, OPP, Fugger and other Gurus have you done good X850XT runs on 3Dmar2005 at near -100ºC?
Any other tips about the X850 problems??

Thanks,
Pedro

BTW : Sunshine day here today, maybe it is time to try again :D

[k|ngp|n]

Too bad about your x850 pedro...why dont you just give up on that thing already, and just go get a x800xt pci ;)

As for pifast, what are you guys scoring under 34sec at 277x13??
Please post your best run.
I believe the wr for pifast is 33.23 by memesama atm.
I wonder how much more lower one gig ram goes on dfi...
Here is mine from three months ago under 34sec at -88c and 275x13 with neo2 and 512mb bh5:
http://pic8.picturetrail.com/VOL256/1324330/2517856/80687093.jpg
I have been wondering how much lower the score can go with 1 gig ram on dfi...

[Crankster]

You pifast better than i PI, where is the justice in the world!
AMAZING job there! :toast: to portugals overclockers!

[mikead_99]

Too bad about your x850 pedro...why dont you just give up on that thing already, and just go get a x800xt pci ;)


Might be the thing to do. FWIW, my run with just -35C wasn't a cakewalk with the X850. Anything above 1.7V would lock my card up too, if the clocks were up there. This wasn't an OCP lock-up. The gpu wouldn't go through a clean reset, like I've seen with X800's, or even 9800's, when OCP kicks in, but it would still crap out. I ended up just finding the happy medium between cooling and volts and living with it. I ended up with the impression that despite the OCP setting, the card's PWM just couldn't handle everything I wanted from it. I'm still hoping to be proven wrong there. I hope you guys do get this sorted one way or the other, the cpu clocks you are putting up are just incredible.

[IvanAndreevich]

This is insane and I hope your get your X850XT PE working OK.

[don_vercetti]

tried an asus x850 xt pe? maybe they are more tolerant of low temps, same as in the 6800u world.

[zakelwe]

I'd go back to a X800XT PE like k|ingP|n says just to stop yourselves going mad. I heard X850 had low nature as well, or was that fixed ?


Almost there though, if it was easy it would not be satisfying to do.

Regards

Andy

[detonator]

I'd go back to a X800XT PE like k|ingP|n says just to stop yourselves going mad. I heard X850 had low nature as well, or was that fixed ?


Almost there though, if it was easy it would not be satisfying to do.

Regards

Andy
bad thing, I dunno where to find an 800XT pcxspress...

[Umek]

Congratzzzzzzz

bye!

[Pedro Rocha]

tried an asus x850 xt pe? maybe they are more tolerant of low temps, same as in the 6800u world.

My X850XT is an Asus, but they are all the same ATI PCB..

Tonight we going to try again ..

OPP and other must have a X850XT secret to their 3Dmark2005 runs ..


Name OPPAINTER,,DFI NF-4,,**Xtremesystems.org 3DTeam**

Description AMD S939 FX-55@ 266x13= 3459MHz,,, DFI NF-4 Ultra,,, 2x512Mb OCZ VX at 266MHz 2-2-2-5,,, ATI X850XT at 853/696,,, PC ICE Cascade cooling the X850XT,,, FX-55 cooled by my other PC ICE Cascade,,, Arctic Silver Ceramique thermal paste,,, OCZ Adjustable 520 PSU,,, Xtremesystems.org 3Dteam
Date 2005-02- 20
User oppainter@xtremesystems.org
Compare URL » http://service.futuremark.com/compare?3dm05=603761


Can you share with us please Mr. OPPainter ??

Thank you :D

[althes]

Good going pedro, I would like to see the x800xt runs though. x850xt not seem to
be cutting it.

[Pedro Rocha]

Good going pedro, I would like to see the x800xt runs though. x850xt not seem to
be cutting it.


If OPPainter and others can have good OCs with their X850, I'm sure we also will :)

As someone of you know, there is a very fair competition and mutual assistance posture between XS members :banana:

[Cpt Twitchy]

Keep up the good work :up:

Hopefully you'll get that X850XT PE running right so you can tear up the ORB.

[gclg2000]

Pedro....that sig is naughty....

[zakelwe]

If OPPainter and others can have good OCs with their X850, I'm sure we also will :)

As someone of you know, there is a very fair competition and mutual assistance posture between XS members :banana:

PM him Pedro, afterall you were kind enough to share the "Ricky tweak"

Regards

Andy

[|RickY|]

well after long night testing the system... i think i finaly can bench the ati, but now i need to speep, its 8:30 am :S all night arround the rig... this is the most intable rig i'v used...

sometimes run lobby@ 3.65 sometimes reboot on windows @ 3.5ghz

tomorrow maybe more results...

FOR BIG PROBLEMS, BIG RESOLUTION :D 220v moded vapochill pe on that :banana::banana::banana::banana:ing mosfets
http://www.teampuss.com/forums/files/vapo_134.jpg

[reject]

sweet! :toast:
are you going to insulate the suction hose on the vapo PE
that ethene makes me REAALY jealous

[DarkManX_BG]

:slobber:
Vapochill on the poor MOSFETS.... if that ain't Xtreme, dunno what is :D

[Paxi]

:slobber: thats just insane :!

[$0m#0n#]

lol, PE on mofsets :D I 'd b happy for PE on my CPU :)

[Pedro Rocha]

Yehh :D , finally we find the solution, 2 mosfest getting really hot even with the card at -100ºC - so for big problems, big solutions ..

Thank you for point us in the right direction :toast:

Now we have to concentrate on the system stability, if we can maintain all test near 3.650 I think we can have one landmark score to share with all of you :D

Ricky's resistance to tiredness and persistence is really amazing, tiredness came erliear to me, but than again I'm almost 19 year older than Ricky :D

Seems we need to do this by turns, but we will never give up :D

As everithing in life - "no pain, no gain" :toast:

[k|ngp|n]

Yehh :D , finally we find the solution, 2 mosfest getting really hot even with the card at -100ºC - so for big problems, big solutions ..

Thank you for point us in the right direction :toast:

Now we have to concentrate on the system stability, if we can maintain all test near 3.650 I think we can have one landmark score to share with all of you :D

Ricky's resistance to tiredness and persistence is really amazing, tiredness came erliear to me, but than again I'm almost 19 year older than Ricky :D

Seems we need to do this by turns, but we will never give up :D

As everithing in life - "no pain, no gain" :toast:

Nice determination guys :toast:
LOL@ the vapo :) ...
Looking forward to some 241+ lh and 198+ ch

[Malachia]

thx for sharing pedro, i start now to build a mosfet cooler....i have here a little 1/12 compressor that seem t be built exactly for that :D

[[XC] leviathan18]

come on guys break the 4ghz... we know that you can... i think you have the "one" chip...

[Malachia]

maybe if you try with LN2 a screen at 4ghz can be near...

[OPPAINTER]

If OPPainter and others can have good OCs with their X850, I'm sure we also will :)

As someone of you know, there is a very fair competition and mutual assistance posture between XS members :banana:
Pedro,

No speacil methods for my card. All of my cards have worked with -90.
What are you useing for the OVP mod?
I suggest no resister, instead of resister just slap a blob of solder over the Cards resister, this will take the OVP pin straight to ground and dissable it.

OPP

[|RickY|]

hi opp, the ovp is working good without removing the resistor

[saaya]

status? :D

[[XC] leviathan18]

saaya can give some advise to have an sli x850xt pe xD

[OPPAINTER]

hi opp, the ovp is working good without removing the resistor
Then just get yourself a XT800PE card. The 800 cards put out higher scores clock for clock against the 850, in 3D2001 anyway.

OPP

[saaya]

same in 2k3 and 2k3 from what ive seen recently iirc... :/

[Crankster]

same in 2k3 and 2k3 from what ive seen recently iirc... :/In both 2k3 and 2k3? Amazing, who wudda thunk it?

[detonator]

cooling down the resistors dident help here.....
saran wrap dient work....the card was dry after 2 double session of 4 houars each ..
a box sealed fealed with co2 aswell(to kill totaly all the condention falled aswell.) we dont have any more hopes with this card, we will swap to a 800XT but we dont know where to find those card anymore.... any tips?
tnx alot.

[k|ngp|n]

Check ebay man....last time I looked there was a gang of them there.

[|RickY|]

try to get bad contact on evap to gpu ;) this will low the temps and maybe help

try to cooler the 3 or 4 chips the card have on back of my vapo cooled place, they get to 60ºc on secounds on 3d run

[detonator]

Check ebay man....last time I looked there was a gang of them there.
your dog own(mastico corso;)

[detonator]

try to get bad contact on evap to gpu ;) this will low the temps and maybe help

try to cooler the 3 or 4 chips the card have on back of my vapo cooled place, they get to 60ºc on secounds on 3d run
I will try ,tnx alot m8:)

[THunDA]

Nice work guys.. man thats some xtreme clocking.. :)

[wfarid]

I'm sure this has already run through ur minds miny times, but subemerging ur board in some cheap non conductive liquid and then phase and run benchies makes sense if ur trying to push ur sys to aneven a deeper and darker oc from where it was b4, u won't have to worry about condensation, the liquid will get cold and cool the board, and keep the mossies under control. Maybe that will be enough for you to get 4gig :)

Great job so far guys!

[Cpt Twitchy]

Rest up so your in your A game and then unleash the beast and show us what that amazing rig can do. I think you have what it takes to get top three in 3D01 if not first.

[M.Beier]

GAAH - I wanna see some SuperPI 1MB :D - Cant wait to see it - AMD has to beat the Intel record at 21 sek !!! :D

Wanna see AMD go for the 19 !!! :D

[Nubius]

Damn nice indeed

[krille]

:thumbsup:

With this it seems quite obvious we'll finally see 4Ghz+ from the E4 San Diegos.

Cheers! :toast:

[GazC]

I'm sure this has already run through ur minds miny times, but subemerging ur board in some cheap non conductive liquid and then phase and run benchies makes sense if ur trying to push ur sys to aneven a deeper and darker oc from where it was b4, u won't have to worry about condensation, the liquid will get cold and cool the board, and keep the mossies under control. Maybe that will be enough for you to get 4gig :)

Great job so far guys!

From what I've seen others say (Charlie for one). Capacitors have a tendancy to pop if they get too cold.

[SAE]

From what I've seen others say (Charlie for one). Capacitors have a tendancy to pop if they get too cold.
Yep. Also proven by some guys benching outside at extremely low temps... Caps in PSU tend to stop working as well as the electrolytic liquid is freezing :D

/edit. Efficiency suffers bad ;)

[GazC]

Yep. Also proven by some guys benching outside at extremely low temps... Caps in PSU tend to stop working as well as the electrolytic liquid is freezing :D

/edit. Efficiency suffers bad ;)

Thanks Stefan, I think you just explained to me why my rig was shutting itself down when I was trying to bench in the garage last month :)

But weird, considering I used to Aircondition my computer!

anyway, I digress, this isn't my thread!

[potkit]

wow man amazing clocks!!! go for the 1st place in the ORB u can do it!!

[gloatlizard]

well sell me your cascade i will pay you 100€ :D :lol:

[LRV]

Yoa are maniacs :D

[M3taC0rT3X]

well sell me your cascade i will pay you 100€ :D :lol:


For the tubbing and insulation ? :lol:


Status ? :P

[Andrewv]

Wow

[Supertim0r]

cooling down the resistors dident help here.....
saran wrap dient work....the card was dry after 2 double session of 4 houars each ..
a box sealed fealed with co2 aswell(to kill totaly all the condention falled aswell.) we dont have any more hopes with this card, we will swap to a 800XT but we dont know where to find those card anymore.... any tips?
tnx alot.

They are still in stock from my supplier:

ASUS EAX800XT-PE 2DHTV 256M Radeon X800XT-PE PCI-E 256MB DDR3 2 DVI+TV-Out HDTV

ASUS EAX800XT 2DT Radeon X800 XT PCI-E x16 256MB DDR 2 DVI+TV-Out Retail

Sapphire ATI X800XT 256M DDR TV DVI VIVO PCI-E Non-Plantium Bulk

[Supertim0r]

saaya can give some advise to have an sli x850xt pe xD
ROFLROFL!!! :p:

[Mag Master 21]

Yep. Also proven by some guys benching outside at extremely low temps... Caps in PSU tend to stop working as well as the electrolytic liquid is freezing :D

/edit. Efficiency suffers bad ;)

Yeah, my OCZ PS 520 refused to post when I left it in 20F temps for 3 hours.

[Gaspar]

Nice Job Boys. Mt folks are from Portugal.. Hence the username.. Viva maritimo.. :rehab:

[Pedro Rocha]

Some updates :

1st the good news :
New WR's on LH, LL and CH .. run with the FX at more than 3.600 and mem between 273 and 277 1.5-2-2-2:D

And now the bad news:
My X850XT are really acting strange - weired results ..sometimes go near 850 core, some other time don't even run at stock ..I suspect it is condensation related :(

I wish we have a tweak to avoid condensation :(

OVP is working - next we gonna try more mosfets cooling ..even with the core at -91ºC (as seen on ATTIToll), we measured some vid card mosfets goin near 60ºC (positive) - that is just amazing :upset:

Seems the same happend to Fugger on Dallas, when our Boss try some extreme ATI OC with phase ..:(

Awful weather here - lots of moisture in the air, bench garage very large and no anti-humidifier there , I wish we have dry night soon :(

[S&M]

Great to hear that !

Can't wait for further updates .... just like the rest of the ppl here at XS ...

[twilius_basic]

Keep on trying guys. It's gonna be worth it, come payday :D

[GuGaCoSa]

I think a LOT of air flow thru the x850 on both sides could avoid the condensation problem,the saran wrap isnt helping that much?

[IvanAndreevich]

Pedro Rocha
Why don't you blow warm air on the card with something like 2x120mm Delta 220CFM fans and see if that helps a bit? :stick:

[[XC] moddolicous]

How are u guys gonna cool the mosfets. I dont know if traditional heatsinks will do it. Good luck though.

[nebuchanezzar]

Nice work :toast: . Have you considered the earlier idea of non-conductive liquid immersion? I'm not on that level of cooling but it may help

[Rabbi_NZ]

Great stuff man! Hope to see that 3Dmark record soon

[EMC2]

--- Pedro ---

Very nice, hope you get dry weather soon :toast:

About the FETs... you're pulling a LOT of current with the core that low and fast. Sounds like their RdsOn stinks :( Get me a close up pic.

If you have nothing that required air cooling...No-condensation tweak - box setup in, purge with nitrogen ;)

[Geforce4ti4200]

wow pushing 4GHz screenshot! how high have you gotten it 3dmark stable so far?

[LardArse]

Some updates :

1st the good news :
New WR's on LH, LL and CH .. run with the FX at more than 3.600 and mem between 273 and 277 1.5-2-2-2:D

And now the bad news:
My X850XT are really acting strange - weired results ..sometimes go near 850 core, some other time don't even run at stock ..I suspect it is condensation related :(

I wish we have a tweak to avoid condensation :(

OVP is working - next we gonna try more mosfets cooling ..even with the core at -91ºC (as seen on ATTIToll), we measured some vid card mosfets goin near 60ºC (positive) - that is just amazing :upset:

Seems the same happend to Fugger on Dallas, when our Boss try some extreme ATI OC with phase ..:(

Awful weather here - lots of moisture in the air, bench garage very large and no anti-humidifier there , I wish we have dry night soon :(

hi pedro my x850 semms to be acting strange too, atitool says -90c in windows at 1.87v. did u monnitor temp with atitool though? when i come out from benching it says -50+c... but drops back to -90c in no time... dont think i have bad mounting but 40c between load and idle is big.

[IvanAndreevich]

I don't see how it's possible to get 40C difference between load and idle if the mounting's good. Did you try with external sensor?

[LardArse]

If the mounting is bad, the pcb won't get as cold as I've felt it :)

[LardArse]

Too late for that all the insulation's up heh.
But i just wanna know if its a temp monitor thingy...if anyone did check the atitool...
either that or my evap's a POS :D

[DarkManX_BG]

Would it be possible to replace the mosfets with better ones? Might be a crazy idea, but I think it's worth looking into it :)

[k|ngp|n]

Some updates :

1st the good news :
New WR's on LH, LL and CH .. run with the FX at more than 3.600 and mem between 273 and 277 1.5-2-2-2:D

And now the bad news:
My X850XT are really acting strange - weired results ..sometimes go near 850 core, some other time don't even run at stock ..I suspect it is condensation related :(

I wish we have a tweak to avoid condensation :(

OVP is working - next we gonna try more mosfets cooling ..even with the core at -91ºC (as seen on ATTIToll), we measured some vid card mosfets goin near 60ºC (positive) - that is just amazing :upset:

Seems the same happend to Fugger on Dallas, when our Boss try some extreme ATI OC with phase ..:(

Awful weather here - lots of moisture in the air, bench garage very large and no anti-humidifier there , I wish we have dry night soon :(

Good to see this thread isnt dead yet :) Keep at it guys.
What about that new pifast wr over 3600?? Lets see those mid 33's...
it only takes seconds to rip one off....no x850 involved.

[Pedro Rocha]

messure the temp's on the back of the card.see if it is realy droping that 40 degree's.

ATITool is not acurate at all, even with stock cooling ..

But my main concerne are the real temperature of the 2 mosfets in the top of the card, measured with a thermometer I see more +40ºC, even with PCB near the core beeing at near -100ºC ..

Motherboard and Vid Card cleaned and isolated again, we gonna try again tonight I hope. :cool:

Just one question: can this be caused by the OCZ Powerstream 600W - that have the 12V splited by 2 rails, thus not enought power on 12v line to feed the X850XT ???

[IvanAndreevich]

Should be more than enough for 1 X850. I remember OPB had 2 Ultras @ 490/1300 on a 520W OCZ.

[Pedro Rocha]

Good to see this thread isnt dead yet :) Keep at it guys.
What about that new pifast wr over 3600?? Lets see those mid 33's...
it only takes seconds to rip one off....no x850 involved.

Will do, of course

We managed to run the mem at 275 1,5-2-1-4 @ 3.7v in the DFI, I'm sure will help on PIFast.. :D

Whats the current WR???

On http://pifast.hexus.net/pifast.php I still see Tom Holck with 33.96, but I suspect there are some scores not submited there ..

[Pedro Rocha]

Should be more than enough for 1 X850. I remember OPB had 2 Ultras @ 490/1300 on a 520W OCZ.

The 600W OCZ have the 12v line splited in 2, the 520W don't - that's the issue ..

[SIOUX]

The 600W OCZ have the 12v line splited in 2, the 520W don't - that's the issue ..


How does the PSU divide the two +12v rails, since you think it could be an issue Pedro ?????


Cant wait to see some results :)

[IvanAndreevich]

I am aware of the fact. But single rail on the 520W was feeding his FX-55 and both of the Ultras. Each of those Ultras is probably like the X850.. What is the AMPs on each of those two rails?

Maybe you can try not o/c'ing the CPU and switching the rails around (one power CPU now to power videocard) and see if anything changes.

[dippyskoodlez]

The 600W OCZ have the 12v line splited in 2, the 520W don't - that's the issue ..

I believe the split rails allows for the cpu rail to pull extra power from the 2nd rail if necessary, but the other rail cant pull back? :stick:

[sandybeach]

He's saying the cpu overloads the 1st 12v rail. Lets say he has 18a and 18a the x850 is only pulling say 12a from rail #2, but the fx55 is pulling 22a from the 18a max on the #1 rail. Since its not a shared rail as on the 520w 40a x1, but 20ax2 it clips OVP..

I guess..

what about this:
Could it be that the reason for these issues is the "RESISTANCE" change that takes place at such low temp's causing more current that can be handled by the circuits to come into play?

Resistance changes with tempratures thus the colder the circuit the more current is pulled for the same voltage. "think i got that right"

maybe this is why we are seeing issues with newer process gpu's and cpu's maybe this resistance change is over powering board power requirments or the parts themselves..

[HaLDoL]

Will do, of course

We managed to run the mem at 275 1,5-2-1-4 @ 3.7v in the DFI, I'm sure will help on PIFast.. :D

Whats the current WR???

On http://pifast.hexus.net/pifast.php I still see Tom Holck with 33.96, but I suspect there are some scores not submited there ..
http://www.ministry-oc.net
record is 33.73 sec

[L0$t Pr0PhEt]

I believe the wr for pifast is 33.23 by memesama atm.


33.73 is kingpins record.

[isp]

kingpin did that a while ago @ 3580 or so

memesama has a 33.23 run for sure that I know of, there might be faster though

[k|ngp|n]

Will do, of course

We managed to run the mem at 275 1,5-2-1-4 @ 3.7v in the DFI, I'm sure will help on PIFast.. :D

Whats the current WR???

On http://pifast.hexus.net/pifast.php I still see Tom Holck with 33.96, but I suspect there are some scores not submited there ..

The hexus site hasnt been updated in along time.
I believe this is the current WR pifast guys...
MEMESAMA:
http://blue.ap.teacup.com/memesama3939/img/1106065915.jpg

[Pedro Rocha]

He's saying the cpu overloads the 1st 12v rail. Lets say he has 18a and 18a the x850 is only pulling say 12a from rail #2, but the fx55 is pulling 22a from the 18a max on the #1 rail. Since its not a shared rail as on the 520w 40a x1, but 20ax2 it clips OVP..
..[/i]

Yes, that's it, thanks for your clear explanation.

Not saying this is the cause, just wondering IF this can be the cause of it..

To bad I RMA my OCZ 520W, and get this 600W back :(

The othery theory - circuits RESISTANCE caused by extreme cooling also makes some sense to me ..

Thank for your imput sandybeach :toast:

[Pedro Rocha]

The hexus site hasnt been updated in along time.
I believe this is the current WR pifast guys...
MEMESAMA:


Thank you :D

So less than 33.23 sec is our target ;), dificult but not totally impossible with this CPU :)

[k|ngp|n]

You guys can do it. With the better mem performance of 1gig on the dfi, ricky & pedro tweak#2 :),high htt with your mem speed@ 275+ tight as nails on div.....low 33's is looking real good :slobber:
Let er' rip man.

[Pejsen]

Funny to see how important tweaking is..

From that ministry-list it is clear that only Tom H benched Pifast above 3700MHz, and yet "only" got 33.96. Memesama @ 3675 gets 33.23.

Ofcourse they're using different ram, but still a big difference i think.

[blinky]

Funny to see how important tweaking is..

From that ministry-list it is clear that only Tom H benched Pifast above 3700MHz, and yet "only" got 33.96. Memesama @ 3675 gets 33.23.

Ofcourse they're using different ram, but still a big difference i think.this happens with all of tom holcks benchies, he runs em at real high mhz, but when other people match his clocks they score way better than him

i think its because of a lack of mem speed + tweaking

[detonator]

Yes, that's it, thanks for your clear explanation.

Not saying this is the cause, just wondering IF this can be the cause of it..

To bad I RMA my OCZ 520W, and get this 600W back :(

The othery theory - circuits RESISTANCE caused by extreme cooling also makes some sense to me ..

Thank for your imput sandybeach :toast:
I have teh OCZ 600 aswell :eek: you make me worryng....

[Pedro Rocha]

I have teh OCZ 600 aswell :eek: you make me worryng....

Ciao detonator :D , do you have any other 24-pin PSU to test ??

[sandybeach]

Opp, had a 520w OCZ If I recall ;)

[Pedro Rocha]

Opp, had a 520w OCZ If I recall ;)

Yes, and OPP have no probs with their ATI X850XT at 853/696. :confused:



Name OPPAINTER,,DFI NF-4,,**Xtremesystems.org 3DTeam**

Description AMD S939 FX-55@ 266x13= 3459MHz,,, DFI NF-4 Ultra,,, 2x512Mb OCZ VX at 266MHz 2-2-2-5,,, ATI X850XT at 853/696,,, PC ICE Cascade cooling the X850XT,,, FX-55 cooled by my other PC ICE Cascade,,, Arctic Silver Ceramique thermal paste,,, OCZ Adjustable 520 PSU,,, Xtremesystems.org 3Dteam

Date 2005-02-20

User oppainter@xtremesystems.org

Compare URL » http://service.futuremark.com/compare?3dm05=603761

[detonator]

Ciao detonator :D , do you have any other 24-pin PSU to test ??
ciao pedro, I made all the test just with this PSU.......
did you try to put the card under vapochill ? maybe is the exstreme cold problem?

[k|ngp|n]

*mICKEYmOUSE* "TeamPuss.com 3DMark Team"

Description A64 FX55-939pin@3375mhz(260x13@1.82v),Dfi Nf4, 2x512 pc3200 OCZ VX at 2,2,2,5,7,12@3.5v, Sapphire X850XT PE@8xx/7xx ;) "gpu@2.09v/ vdd@2.18v/ vddq@2.22v" "cpu dryice cooled / gpu swiftech water cooled with chilled water at -30c "OCZ 600w Power supply"

No problems there.....

[Pedro Rocha]

ciao pedro, I made all the test just with this PSU.......
did you try to put the card under vapochill ? maybe is the exstreme cold problem?

Sorry ..do you mean the Vid Card under Vapochill ??

The CPU and mobo yes, the X850XT allways run on stock cooler before going to cascade ..

[Pedro Rocha]

*mICKEYmOUSE* "TeamPuss.com 3DMark Team"
Description A64 FX55-939pin@3375mhz(260x13@1.82v),Dfi Nf4, 2x512 pc3200 OCZ VX at 2,2,2,5,7,12@3.5v, Sapphire X850XT PE@8xx/7xx ;) "gpu@2.09v/ vdd@2.18v/ vddq@2.22v" "cpu dryice cooled / gpu swiftech water cooled with chilled water at -30c [/B] [B][SIZE=5]"OCZ 600w Power supply"

No problems there....



Yeh ..runnig out of ideas to be honest :mad:

[detonator]

yea I mean to try to put the video card under Vapochill@-50 dry icy -70/6°, maybe -95 or more is just too much......

[sandybeach]

I take it you guys have ran the CPU on air cooling and GPU only on cascade?

I would get a brand new DFI NF4 and a 3000+ and on stock air drop your cascadd x850's in and test ;)

Or what about tring a diff motherboard like an MSI Neo4

[[_Silence_]]

I take it you guys have ran the CPU on air cooling and GPU only on cascade?

I would get a brand new DFI NF4 and a 3000+ and on stock air drop your cascadd x850's in and test ;)

Or what about tring a diff motherboard like an MSI Neo4

They're running 2 cascades, 1 for CPU and another for GPU.

[EMC2]

--- Dippyskoodlez ---

Nope, they don't share current at all, seperate OCP for sure and usually seperate regulation on the two rails.

--- Darkman ---

Yes, if that was the problem, they could be.

--- Pedro ---

If the FETs are only getting to 40C, that is not an issue... for a FET that is cool. You could always add a small sink to verify.

Regarding the power issue... have you measured the input rails to the Vid Card under load (12V, 3.3V)? If it's a supply current issue, you should see them sagging. No sag, no issue. Also, if you have a good scope, look at the regulator outputs on the card... maybe too much noise or fluctuation on them with the current draw at low GPU temps.

[Pedro Rocha]

--- Pedro ---

If the FETs are only getting to 40C, that is not an issue... for a FET that is cool. You could always add a small sink to verify.


Not and issue maybe, but anyway quite strange, the rest of vid card PCB is quite cool ..from near -100 on back of GPU to very negative temperatures on rest the of the components - I have and infra-red temp meter and checked ,,

If the card was not sub-zero cooled these components can easy pass +100ºC ..or even explode like happend in Dallas ATI Event..

I have a lot vid cards at sub-zero in the past, and never saw that .. :slobber:

Anyway we put some tweakmonster there ...

Also gonna try the OCZ 520W ..in any case ..

Thanks :D

[IvanAndreevich]

If the problem lies with the temperature of the FETs, it has to do with the temperature DIFFERENCE, not the overall temperature. If they are +60 and the rest of the PCB is -100, there could be some traces which glitch from expanding / contracting.

[sandybeach]

They're running 2 cascades, 1 for CPU and another for GPU.

I know, lol Im saying get a neo4 with a winny on stock air and slap the x850+gpu cascade on that setup ;)

[STEvil]

There are two schools of thought on muliple sources for the same rail.

School 1 states that each rail is seperate from the other, in which case we would see components dieing due to voltage differential generated by excessive current draw (ie: like some x800's did when a second dedicated PSU was attached to them).

School 2 states that there are two transformers inside the PSU pooling their outputs which will eliminate this problem since it is effectively a single rail with two input sources. In essence 20+18 = 38 useable by anything.

School 2 decreases transient response times (good), allows less voltage fluctuation (good), can clean out AC voltage better (good), and can also have a more stable frequency shift (ie: mV variations at 5-20hz, good).

I dont know how many PSU's use 1 and how many use 2, but you get what you pay for..

[Malachia]

yea I mean to try to put the video card under Vapochill@-50 dry icy -70/6°, maybe -95 or more is just too much......


stock also had some problem with my mach 2 GT @410 (temps like dry ice...), i have the fear that ati used a new controller like the one nvidia used, for thermal protection...

[G H Z]

Are the running voltages stable on the card(s) prior to lock-up? Have you tried resetting the overclocking utility to default, closing it, then re-opening it and try again? Maybe it's the card utility itself, might ask around about that.

[Pejsen]

this happens with all of tom holcks benchies, he runs em at real high mhz, but when other people match his clocks they score way better than him

i think its because of a lack of mem speed + tweaking

Yep i agree.

[sandybeach]

stock also had some problem with my mach 2 GT @410 (temps like dry ice...), i have the fear that ati used a new controller like the one nvidia used, for thermal protection...

There is "NO" protection, lol.. How is it that does OPP and others get by the low temp then on there x850xt pe? Are they GOD so they just click there fingures and it all goes away?

The issue is in the complete package not just the card. If you don't see that then......

I still haven't got a reply on the msi neo4 +stock air cooling on a cpu and gpu under cascade??????

[reject]

@Pedro have u tried the reject mod
join both 12 rails to the card
i dont have my camera now, will upload a pic tomorrow

[regedit]

Thats fkn pretty nasty :banana::banana::banana::banana: you got there.WD

[regedit]

cool

[reject]

behold:

http://www.xtremesystems.org/forums/attachment.php?attachmentid=27611&stc=1

[IvanAndreevich]

Sorry what the heck is that?

[dippyskoodlez]

Um... whats so special about a molex adaptor?

[Cpt Twitchy]

Can't you see that it is an uber 1337 platinum plated molex connector. :D

I actually have the same question.

[d3birth]

@Pedro have u tried the reject mod
join both 12 rails to the card
i dont have my camera now, will upload a pic tomorrow


It's a picture of that.

[reject]

yes, pedro was worried about 1 12v rail sharing the load of cpu and gfx, and i have the solution. split the load of gfx onto both rails :D

[STEvil]

To do that your solution would require the rails to be seperated internally and if one component pulls enough amperage to overpower one rail its going to pull through your connector what it needs to make up the difference...

What yours is likely doing is making up for losses at the connector due to low pressure and bad surface contact area.

[Pedro Rocha]

Well, another bench session, anoter AMD WR ..

This time using CPU-Z 1.27:) ..

http://pedrorocha.planetaclix.pt/Fx3899.jpg

We have another screens, but F5 in CPU-z was jamed by so much speed ..the mythical barrerier is near, very near ..but we must fight for the last steps :D

Thank you Ricky, you build 2 fantastic cascades :toast:

[G H Z]

That is very amazing guys, good job :toast:

[Pedro Rocha]

And 3 more ransonable ;) scores with the X850XT with stock cooler / stock speed and and the CPU @ just around 3.600mhz ..

http://pedrorocha.planetaclix.pt/Loby3571.jpg

In the bench that really matter for me and Ricky :D

[[XC] moddolicous]

wow, you are soo close to 3.9! I cant beleive u didnt reach 3.9 yet. Did u guys get the X850 fixed? Once u get 3.9, u might wanna start a new thread.

[|RickY|]

help me ppl, help-me to convence Pedro to leave-me get that :banana::banana::banana::banana:ing HS away :D :D

i know that cpu isn´t for broke :) i'm kidding :D

[Pedro Rocha]

wow, you are soo close to 3.9! I cant beleive u didnt reach 3.9 yet. Did u guys get the X850 fixed? Once u get 3.9, u might wanna start a new thread.

We did reached :D , but tonight the hunt was for the Big 4 :D

[gclg2000]

I hope you guys get it. I didn't see any info on the cascades on the 1st post of this thread? Are ya'll running anything in a vaccum?

[[XC] moddolicous]

Did u take the IHS off the CPU. That might allow you to reach lower temps, then allowing u to reach 4ghz.

[cpulloverclock]

edit your avatar :)

where are the benchmarks? :D

cpumark, superpi, pifast, sandra sisoft cpu scores, cinebench.... :slobber:

[SLaY3r07]

Hopefully 4GHZ will be here soon...

[cpulloverclock]

Did u take the IHS off the CPU. That might allow you to reach lower temps, then allowing u to reach 4ghz.
dangerous with a dod, chill, cascade :shock2:

[[XC] moddolicous]

dangerous with a dod, chill, cascade :shock2:
what do u mean. I know it would be dangerous, but so is overclocking a stock 2.6ghz chip up to 3.9. What do u mean DOD.

[cpulloverclock]

what do u mean. I know it would be dangerous, but so is overclocking a stock 2.6ghz chip up to 3.9. What do u mean DOD.
direct on die

as he feels it

[Cpt Twitchy]

Keep up the mad oc'ing skills.
Congrats on all the world records.

[Tony420]

NICE work. I wonder how far this super OC goes with prime stability. probly like 3.6! (which is insane)

[reject]

To do that your solution would require the rails to be seperated internally and if one component pulls enough amperage to overpower one rail its going to pull through your connector what it needs to make up the difference...

What yours is likely doing is making up for losses at the connector due to low pressure and bad surface contact area.
yes, my psu has 2 separate 12v rails, separated internally. so does perdros from what i gather

[STEvil]

What PSU? Enermax?

I have raised the question on the 600w Powerstreams and have been told they are connected internally, but i'm still trying to get info on if its just a signal connection to minimalize voltage differential or a full blown heavy duty PCB trace that can handle some amperage...

[LardArse]

Hey there Ricky and Pedro, good luck on that 4G. :toast:
Have you gotten the X850 to run properly at cascade temps yet? I'm having a bit of problem with mine too if its any assurance :)
How does your card act up? Freezes, reboots or the multiline splash screen syndrome?
Does feeezing the hot fets help? :)

[reject]

yep, noisetaker 470w. 12v1 is 16A and 12v2 is 15A so im sure it helps

[sandybeach]

Look no offense to that awsome cpu clocking, but why not "FINISH" a damm air run guys? hehe

Thats 42k on air. Thats higher than your old 40,600 score and if you fail to never get a full bench with x850 on cascade you have one of the top 10 scores in the world on air only?

Also any news on the gpu on cascade in a neo4 with a cpu on air cooling yet?