Sandy Bridge 5GHz + Club

[OC Nub]

OC Nub, 1M is pure raw CPU power... 32M should be beneficial of tighter latencies and more ram speed

Im slowly learning this lol.

My luck with good sb chips hasnt been the greatest. I hate to buy anymore chips in hopes of a super chip, probably best to stick with what I have and wait for next platform.

[TWK_OCZ]

i5 2500K @ 5.30 GHz - 1.600 V.- Asus P8P67 PRO - G.Skill XM - WATER - #L041

[l0ud_sil3nc3]

Great buddy

my UD7 should be here on Wednesday can't wait

[andressergio]

my UD7 should be here on Wednesday can't wait

best of lucks mate

[l0ud_sil3nc3]

Im slowly learning this lol.

My luck with good sb chips hasnt been the greatest. I hate to buy anymore chips in hopes of a super chip, probably best to stick with what I have and wait for next platform.

no kidding about the bad luck on sb, I have been through so many 52x chips and some 53x but thats it, and I only have one good 55x chip that came from a member on here.

[newhit]

no kidding about the bad luck on sb, I have been through so many 52x chips and some 53x but thats it, and I only have one good 55x chip that came from a member on here.

Sorry for numpty question but............do you mean a chip that will get into "Windows" at 55x and 52x and bench or one that will just post.

I'm getting confused by the various threads regarding stability v raw power........... and Blend v Small ssts v LinX v IBT v SuperPi etc.

If somebody's chip does 5.2@1.41v and bench, is it better than a chip that will do 5.4 but require silly Vcore (i.e. mine)?

My L045B398 will run Blend for 4 hours at 5GHz 1.44v and do LinX at higher multi's for 20+ loops with correspondingly higher Vcore (I know its not the best but not the worst either) but I switched off because I wanted to try another chip.

Where does the fun end and frustration kick in, so to speak.......

[Leeghoofd]

If it enters Windows XP ( far harder stress then Vista and Win7) then you have a good one... when we did the binning of +50 CPU's it took us about 4-5 minutes per CPU do determine approx speed... if it posts into windows at 55X with eg 1.55Vcore then it will run +5500 for benching... we found 3 X 56 multies : one does 5730, one 5700 and the worst one 5660....

Posting is not good enough... and make sure when you test it has EIST and co all off...

[l0ud_sil3nc3]

If it enters Windows XP ( far harder stress then Vista and Win7) then you have a good one... when we did the binning of +50 CPU's it took us about 4-5 minutes per CPU do determine approx speed... if it posts into windows at 55X with eg 1.55Vcore then it will run +5500 for benching... we found 3 X 56 multies : one does 5730, one 5700 and the worst one 5660....

Posting is not good enough... and make sure when you test it has EIST and co all off...

@ newhit

exactly what Leeghoofd stated, boot in to windows and bench, i.e. superpi 32m, pifast, etc. . .

most people that are binning are looking for 56x and greater chips and it seems to be very hard to find one

[newhit]

l0ud_sil3nc3 and leeghoofd, many thanx.

Just one more question. If you could answer the following......

I have win7 (SP1 with AVX). If my chips need 1.55+ Vcore to bench can I not still claim stability, even if I'm nearing 1.6v Vcore (Iam on water). Is it not tough enough within the Win7 (SP1+AVX) benching environment to claim "reasonable" stability. Obviously my temps are going to be the wrong side of 85 C and the "degradation" word comes in but for arguments sake is this amount of abuse still not a good claim?

For the sake of practicality and the "fun" word, an hour or two of Prime must be good to go (or a good 20+ loop in LinX maybe, or IBT or..........etc).

Once again thanx for the replies

[Charloz24]

Don't know if it's superpi 1M stable, but here is my 5ghz shot



Uploaded with ImageShack.us

[coolhandluke41]

how about 32M @5389 ;

here is 1M;


Uploaded with ImageShack.us

[OC Nub]

l0ud_sil3nc3 and leeghoofd, many thanx.

Just one more question. If you could answer the following......

I have win7 (SP1 with AVX). If my chips need 1.55+ Vcore to bench can I not still claim stability, even if I'm nearing 1.6v Vcore (Iam on water). Is it not tough enough within the Win7 (SP1+AVX) benching environment to claim "reasonable" stability. Obviously my temps are going to be the wrong side of 85 C and the "degradation" word comes in but for arguments sake is this amount of abuse still not a good claim?

For the sake of practicality and the "fun" word, an hour or two of Prime must be good to go (or a good 20+ loop in LinX maybe, or IBT or..........etc).

Once again thanx for the replies

I dont think anyone trying for x55 or higher is looking for anything but bench stable, meaning it will pass the test and stay in windows long enough that they can take a screen. Lower voltage chips are nice but there are no points taken away for high vcore, you just need to get the screenshot.

[chew*]

1m 5 gig + runs seems kinda pointless.

5 gig + stress test stable would make things more interesting.

Chip has a lot more headroom for stress testing, awsome low volt chip.

[newhit]

I dont think anyone trying for x55 or higher is looking for anything but bench stable, meaning it will pass the test and stay in windows long enough that they can take a screen. Lower voltage chips are nice but there are no points taken away for high vcore, you just need to get the screenshot.

Cheers "OC Nub".

You seem to be less than happy with your SB experience. What is the reason for this? "Stability" at 5GHz+ with low Vcore?

[ChaosAD]

Indded having true stability at 5Ghz @ 1.34v is very nice. But with my bad luck is impossible to find a similar chip.

[OC Nub]

Cheers "OC Nub".

You seem to be less than happy with your SB experience. What is the reason for this? "Stability" at 5GHz+ with low Vcore?

I am happy with SB. I havent tried stability at 5ghz, I have spent most of my time 3d benching at 5.4ghz.

[PolRoger]

1m 5 gig + runs seems kinda pointless.

5 gig + stress test stable would make things more interesting.

Chip has a lot more headroom for stress testing, awsome low volt chip.

Nice one chew*!!!!

Now that is the kind of 2600K I've been looking for... that one should be able to do 48x with ~1.3v or less. 55x(+)?? Is that one of your better ....C108's?

[newhit]

I am happy with SB. I havent tried stability at 5ghz, I have spent most of my time 3d benching at 5.4ghz.

Note to self........must update psychology to "xtreme sys" level.

[chew*]

Nice one chew*!!!!

Now that is the kind of 2600K I've been looking for... that one should be able to do 48x with ~1.3v or less. 55x(+)?? Is that one of your better ....C108's?

Scaling roger appears to be .025 per 100 mhz up to 5.4 after that .050 per 100mhz to 5600 ish. I only tested prime to 5.4 but cut it at 1 hour for safeties sake.

Yes C108, they seem to do great with low volts, even cold once maxxed more volts pisses them off.

[andressergio]

1m 5 gig + runs seems kinda pointless.

5 gig + stress test stable would make things more interesting.

Chip has a lot more headroom for stress testing, awsome low volt chip.

very nice chip chew