XS HWBot Team

[mk-ultra]

already outdated, 117k @ 03 , a few more runs on water to come

[Linuxfan]

already outdated, 117k @ 03 , a few more runs on water to come

I will definitely stick my GTX 480 on LN2 next time I bench it with my new CPU this time.... Big gains in 03 with GPU clock apparently.

[Linuxfan]

3870X2 results improved with GPU overclocked higher this time and new CPU... More tweaking to do later then LN2 on CPU.

[luke997]

Wow, nice scores you guys rock

Any ideas why Wprime and Aquamark would not start on my rig?
Wprime crash on startup with error (without starting any calculation) and Aquamark just would not start (tried admin mode, no joy).

Got 3d 2003 & 2005 now so will be running shortly

[Leeghoofd]

Wprime run as administrator on Vista/win7 (should work unless very unstable) Aquamark3 needs a 64 bit patch

[luke997]

Wprime run as administrator on Vista/win7 (should work unless very unstable) Aquamark3 needs a 64 bit patch

Thanks, both worked fine, got a crap score but every penny counts

[luke997]

Guys, sorry for double post - forgot to ask about it earlier.

Any ideas how what vcore for i7 980X would cause it to degrade - when applied in couple hours benching sessions (mostly Futuremarks)?

Example:
Vcore: 1.56v (as reported in CPU-Z under load, done through offset so dropping to ~1.27 under light load)
QPI: 1.35
DRAM: 1.64(or 1.66)
PLL: 1.90

With LLC on at 1000mv, on water temps well under the safe margin.

Any thoughts? I realise that every cpu is different etc. but I reckon who else can have some first hand experience

[Linuxfan]

Guys, sorry for double post - forgot to ask about it earlier.

Any ideas how what vcore for i7 980X would cause it to degrade - when applied in couple hours benching sessions (mostly Futuremarks)?

Example:
Vcore: 1.56v (as reported in CPU-Z under load, done through offset so dropping to ~1.27 under light load)
QPI: 1.35
DRAM: 1.64(or 1.66)
PLL: 1.90

With LLC on at 1000mv, on water temps well under the safe margin.

Any thoughts? I realise that every cpu is different etc. but I reckon who else can have some first hand experience

I know I and others who have had 45nm i7s like the 920s who ran them on phase or cascade at 5ghz high voltages through futuremark benches for over a year all the time. Sometimes after a while the chip will wear down and you'll have to lower the clocks a little bit(maybe 50-100mhz) with the same voltages to get stable again. I don't think its bad though.
If you beat it hard it for 5 years though it would probably degrade a lot. But considering, most people upgrade before then and that will not be there top end benching system anymore.

[luke997]

I know I and others who have had 45nm i7s like the 920s who ran them on phase or cascade at 5ghz high voltages through futuremark benches for over a year all the time. Sometimes after a while the chip will wear down and you'll have to lower the clocks a little bit(maybe 50-100mhz) with the same voltages to get stable again. I don't think its bad though.
If you beat it hard it for 5 years though it would probably degrade a lot. But considering, most people upgrade before then and that will not be there top end benching system anymore.

Thanks for reassurance

[Linuxfan]

More cards incoming....
7950GT and 3870 on the way!

[Don_Dan]

More cards incoming....
7950GT and 3870 on the way!

Sounds great! Will you go for some triple CFX numbers as well?

@luke997: Please fix some of your screenshot, they were resized and you can't see anything at all on them. You risk your scores getting blocked.


By the way, what happened? Looks like we lost 60 points again? Did someone leave?

Oh, and very nice scores MSIMAX!

[luke997]

@luke997: Please fix some of your screenshot, they were resized and you can't see anything at all on them. You risk your scores getting blocked.

There's something wrong with editing, tried twice, said all fine, yet no changes - is there a time lag before you can see it? ( I've cleared it from browser so no caching issue).

[Don_Dan]

I know that problem, sometimes changes will not stick, it's hwbot's problem, not a problem of cached info imho.

So your only chance is to run even faster und submit a new score!
( One more thing to note, don't update scores when you hit a higher score, just submit a new entry, hwbot doesn't like updating scores. )

[Linuxfan]

[Don_Dan]

Please welcome our new member NAMEGT!

[Linuxfan]

Please welcome our new member NAMEGT!

Glad you joined!

Nice AMD CPU-Z record there!

[namegt]

Thanks ~
The lack of my ability, but I'll try for the team.

[MrIcee]

namegt

It is with great pleasure that I say welcome to the XS Team
Looking forward to your future work and results. We are Xtremely
happy to have you aboard. Good luck with your future benching

Best regards,

Randi

[vistageek92]

Im trapped in Linuxfans garage breathing nitrogen fumes.....

Working on top 5, GTX470 benches. Currently have a world record 6th place, but where shooting for a 5th or higher. :smokin

We are having a little trouble with cold bugging, which shouldn't be happening so we are tweaking RAM timings and such at the moment.

Hardware:

980x @ 5.6-5.8ghz
Rampage 3 Extreme
GTX470
6GB mushkin redline

55 liters of liquid Nitrogen.

We had been using both a CPU and GPU pot, but the GPU was being a pain, so we threw it back on air.

Xtremesystems best single card result here I come!!

Pics:

[dafour]

Nice,good luck!

[=[PULSAR]=]

Anyone know where I can find a core 2 duo OC database I want to go shopping around for a E8600.

[thijs]

Q822A is the golden batch i believe. if thats what you mean.

[=[PULSAR]=]

Really!?!? If so I have an opportunity to get that exact batch.

Q822A566

[thijs]

Q822A435 was like the known best, you can google it. not to sure about the 566, and YMMV obviously. google for a bit i would say.

[Don_Dan]

Yes, Q822A435 and Q820A598 are the great batches, in general the early batches ( Q820 and Q822 ) are better than more recent ones.