EVGA X58 Test on i7 965XE

[LardArse]

Benching @ -120C. 3D Mark 06 @ close to 5GHz in Vista:

http://d.imagehost.org/0189/4980.jpg

Seems that 3D Mark 06 reports the wrong cpu speed...
This will happen when you use above default multiplier since higher multipliers is acheived through turbo mode, and that comes in only when OS is loading, that's why probably you will many pictures of insane overclocks at boot up since then it is still on default multiplier no matter what you set :)


3D Mark Vantage Test @ 4.92GHz PhysX disabled of course:

http://d.imagehost.org/0668/nben2.jpg


And also decided to play with the 'Brink OC' function in ELEET Tune :D

http://d.imagehost.org/0654/nben1.jpg

http://d.imagehost.org/0214/nben3.jpg

Maybe some experience no scaling with voltage , I definitely can scale with voltage on this board :)

[Hazzan]

Nice result with new your product Shamino.....:up::up:

[bartx]

I thought that Nehalem performs better in 2k6 CPU Test. But in Vantage cpu score is so nice :D Keep pushing it :up:

[LardArse]

Thanks, selectable PWM Frequency via BIOS :)

http://d.imagehost.org/0719/pwm.jpg

[Benny Lodewijk]

New monster ! Nice result Peter

[artkobe]

good job peter!
looking forward 5Ghz+ 3Ds
and where i can down EVGA E LEET?Is it fits for all MB?

[Gautam]

Excellent speeds! :clap:

[tiborrr]

Great job Peter, looking forward to see more! :up:

[splmann]

Good Job Peter !!!

[SoF]

yeah great to see you playin' wit i7 - nice start, can't wait for some scores with some of your full modded GPUs :D :D :D what vdimm you're using mostly?

[LardArse]

yeah great to see you playin' wit i7 - nice start, can't wait for some scores with some of your full modded GPUs :D :D :D what vdimm you're using mostly?

I used 1.95v but it seems just 1.75v is enough to run 6-5-5-18, but just testing cpu speed first to be assured the pwm works good for the high loads :)

[SoF]

thanks for info, looks like under cold it can take enough vdimm for reasonable memory timings with that divider. cpu looks good so far, hope you can bench 5gig+ soon :D

[makaka]

Nice work what about some PC mark or winrar test i am pretty sure u will step over 6 or even 7k

[Oliver]

thx for posting Peter-tan

the board seems very promising

keep pushing it

[TheKarmakazi]

Thanks, selectable PWM Frequency via BIOS :)

http://d.imagehost.org/0719/pwm.jpg

What are the effects of changing the PWM frequencies?

[Element009]

nice results

We are waiting for more benches^^

[M3kk]

Very nice Peter :).

[HDCHOPPER]

exellent job peter !!

sham da man:up:

[ZoLKoRn]

Sweet board from evga!:up: that really looking interesting board ;)

looking forward for 3d benching :D

[AuDioFreaK39]

wow awesome results Shamino :up:


what kind of memory are you using?

[JacobF]

Check out this result on Air with a 920!

http://valid.x86-secret.com/show_oc.php?id=443907

http://img510.imageshack.us/img510/7987/w00tuf6.png

[T_Flight]

Guys! :eek: I'm am stunned! I knew this thing was awesome, but never dreamed this CPU would be this high up there already so soon. The air shots I'm seeing are incredible. I plan on running mine on water and hate to think what it's gonna do. I'm not quite ready for phase. I might try phase one day, but wanna work with water for awhile, and maybe advance to chilled water.

Intel and all the board manufacturers have really done an oustanding job with this platform. I cannot wait until the 17th.

Thanks so much for posting this info. It's really helpful, and I've learned alot. :)

[JacobF]

What are the effects of changing the PWM frequencies?

Higher frequency is cleaner power :yepp:

[bazx]

good to see you have to m/b rocking peter

very interesting about the pwm

[DeDaL]

Its time to Aquamark/3Dmark 2005 play.

Nice board Sham and nice CPU.

[SF3D]

The board seems to be good! Nice CPU Peter!

So, EVGA jumped to digital PWM too. Good choice!

Can't wait to test this one ;)

[LardArse]

Vantage CPU Test is ok @ 5GHz:

http://d.imagehost.org/0034/5gvanm.jpg

Little LN2 left so tried disabling HT for a quick Pi run:

http://d.imagehost.org/0876/533311.jpg

[LardArse]

What are the effects of changing the PWM frequencies?

the higher switching freq reduces little minute ripples your multimetter won't pick up. your mosfets and inductors are more stressed at higher frequencies so if you want to run cooler mosfets or save energy lower fequency

[K404]

Thanks for the PWM info Peter. What are the limits for the switching frequency- at max selectable switch Freq, is the controller on edge or straying into any gray area?

For some other folk- remember a lot of 7900GTs needed a PWM freq mod for higher GPU core ;) The PWM wasnt keeping up at default freq.

[LardArse]

Thanks for the PWM info Peter. What are the limits for the switching frequency- at max selectable switch Freq, is the controller on edge or straying into any gray area?

no, tested loaded @ 1,200+KHz fine. But throw in a buffer region and so 1.05k....

[K404]

That reply was almost as fast as your SPi1M :p: Thanks!

[LardArse]

That reply was almost as fast as your SPi1M :p: Thanks!

that was a crappy run no doubt heheh no tweaks but :banana::banana::banana::banana: the ln2 consumption during pi and cpu test is so huge :D

Well many ICs you can mod them higher, but you need to monitor each pahse, the msofets may start "overlapping" each other when switching faster so this is no go for higher freq, you need to switch fast but not overlap.

[K404]

Oscilloscopes should be part of everyones lineup, just like thermometers and DMMs ;) :p:

If the LN2 consumption is crazy high for even single-threaded apps, will there be an option of disabling cores as well as HT? (Iv not played with any Nehalem boards, so that might be a dumb question)

:toast:

[LardArse]

Oscilloscopes should be part of everyones lineup, just like thermometers and DMMs ;) :p:

If the LN2 consumption is crazy high for even single-threaded apps, will there be an option of disabling cores as well as HT? (Iv not played with any Nehalem boards, so that might be a dumb question)

:toast:

no i mean single thread eats so little ln2 compared to 8 threads

[massman]

the higher switching freq reduces little minute ripples your multimetter won't pick up. your mosfets and inductors are more stressed at higher frequencies so if you want to run cooler mosfets or save energy lower fequency

Interesting. Does it really work? Higher stable frequencies and saving energy?

[LardArse]

Interesting. Does it really work? Higher stable frequencies and saving energy?

no no lower freq to save energy, increase freq for oc

[massman]

Yes, but does it actually work? Manufacturers claim that their features kick ass all the time, I wonder if it réally works and if yes, on what scale it helps :).

[dinos22]

very intesting board
:o

all the big boys are gonna start paying attention to their new competitor for sure eeek

show us some more

that 920 OC is just :o
what ambient jakup?

[LardArse]

Yes, but does it actually work? Manufacturers claim that their features kick ass all the time, I wonder if it réally works and if yes, on what scale it helps :).

i didnt claim it kicks ass... but you have a choice, itsn't that always good? :)

[T_M]

your mosfets and inductors are more stressed at higher frequencies so if you want to run cooler mosfets or save energy lower fequency

Thx, exactly the answer to the question i was about to ask (Why make it adjustable if higher = better). :up:

[SoF]

wohooo that looks really promising - one more 5gig+ i7 system in good hands :D tried running 1 or 2 cores if that can give you something in 1m?

[WaterFlex]

Peter, well done mate.

[L0$t Pr0PhEt]

Yes, but does it actually work? Manufacturers claim that their features kick ass all the time, I wonder if it réally works and if yes, on what scale it helps :).

Normally it should work, the higher the pwm frequency the better the signal can be made but then you get more heat production. Every time it switches you have a small energyloss so lower frequency should reduce the energyconsumption of the board.

[HDCHOPPER]

sham is this your EVGA board or next gen will have your brain on it ??

eyema sham follower :D everything you touch turns to gold :D

[Solid Snake]

Excellent Score

you have a monster !!!! :clap:

[massman]

Normally it should work, the higher the pwm frequency the better the signal can be made but then you get more heat production. Every time it switches you have a small energyloss so lower frequency should reduce the energyconsumption of the board.

But there's a big difference between should and does work. I was kinda hoping for Peter to give me raw numbers of the performance scaling :p:

[AuDioFreaK39]

sham, are there any trends you've learned on the 965 between raising the BLCK vs. raising the multiplier? Which one is is best to start raising first?

[2chesapeakes]

Very nice results.

[Razzz]

Nice! 53xxMHz is really good.

[scooter.jay]

Nice work shamino some great results coming from i7. More settings i have only just got used to the last lot:rofl:

[hey]

wow look at that evga e-leet utility :D cpu-z visualization :)

[LardArse]

wow look at that evga e-leet utility :D cpu-z visualization :)

Franck is such a nice guy to work with, Jakup and I have thoroughly enjoyed our time co-working with this talented man :)

[Harshal]

Having fun the Xtreme/Sham way :up: