Sandy Bridge and stress testingPost Date: 2011-01-18 |
Post Reply
|
Author | |
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply
Topic: Sandy Bridge and stress testing Posted: 18 Jan 2011 at 6:29pm |
Just a heads up for all you people wanting to overclock Sandy Bridge, also Digital Storm may all ready be aware of this.
Linx or Intel burn test will not stress Sandy Bridge, unless you use Windows SP1. Linx should produce 120-140 G-flops with Sandy Bridge overclocked. Without SP1 windows will not stress Sandy bridge enough, but with SP1 it will need more voltage to run stable. Edited by justin.kerr - 18 Jan 2011 at 6:33pm |
|
DST4ME
DS ELITE Joined: 14 Apr 2008 Online Status: Offline Posts: 36758 |
Quote Reply Posted: 18 Jan 2011 at 9:05pm |
Thx for that info justin
|
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 19 Jan 2011 at 8:00am |
finally got my hands on SP1 final, did a quick test, need to do more, but F'ning amazing. lol.
|
|
Alex
Admin Group Digital Storm Supervisor Joined: 04 Jun 2012 Online Status: Offline Posts: 16314 |
Quote Reply Posted: 19 Jan 2011 at 1:02pm |
Wow nice! I'll have to update to SP1 and try it myself.
Have you tried Prime95 on Blend? That's the real kicker for Sandy Bridge... |
|
tech007
Newbie Joined: 04 Nov 2009 Online Status: Offline Posts: 46 |
Quote Reply Posted: 19 Jan 2011 at 1:22pm |
Justin.Kerr what cooling did you use for that test?
Edited by tech007 - 19 Jan 2011 at 1:22pm |
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 19 Jan 2011 at 2:12pm |
Alex, only tried Linx, but what did pass 20 passes of Linx, exact same settings neded about 4 clicks more voltage to be stable for just 3 passes on SP1, just to let you know
|
|
Alex
Admin Group Digital Storm Supervisor Joined: 04 Jun 2012 Online Status: Offline Posts: 16314 |
Quote Reply Posted: 19 Jan 2011 at 2:53pm |
Thanks for the heads up!
|
|
Metamorphosis
Newbie Joined: 25 Dec 2010 Online Status: Offline Posts: 62 |
Quote Reply Posted: 19 Jan 2011 at 4:37pm |
Prime95 is acting really weird with SandyBridge atm.. if you disable hyperthreading it runs stable, but with hyperthreading on its alot harder to get even 20mins on Prime95.
And Vantage showing lower GPU scores with hyperthreading on then off is also concerning since Vantage as far as I can remember has always favored hyperthreading. Hopefully some patches will resolve the issues. Edited by Metamorphosis - 19 Jan 2011 at 4:38pm |
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 19 Jan 2011 at 8:37pm |
Wow, that is a load BS, more like a train load.. lol
the reason people would be having trouble with HT on, is because it puts more stress on the CPU, and needs more voltage, same with all HT Intel processors. this has been a common fact known for many years. lol Mine runs Prime95 stable for hours at 5,100Ghz with HT on.
and 3dmark Vantage is averaging around 30% better CPU scores with HT on VS HT off.. Just looked at my testing, and the top 30 world records in 3Dmark Vantage with 2600K, all used HT. GPU score is just that, dependent on GPU, CPU score is just that, CPU dependent, unless you are a total noob,and have the GPU do the CPU calc.. lol it happens.
|
|
Metamorphosis
Newbie Joined: 25 Dec 2010 Online Status: Offline Posts: 62 |
Quote Reply Posted: 20 Jan 2011 at 1:05am |
Wow man.. just completely dismiss it hey? Did you actually look? And I'm not talking about CPU score I was talking about GPU score.
HT on HT off More anomalies with the GPU and hyperthreading.. you must have one heck of a golden egg to be the only person not experiencing this. |
|
Dragoonseal
DS Veteran Joined: 03 Apr 2009 Online Status: Offline Posts: 2247 |
Quote Reply Posted: 20 Jan 2011 at 3:57am |
Wait, so you got a 0.88% difference in score on the second run and didn't think that maybe, just maybe, that kind of falls within the margin of random variance territory?
Oh that perhaps Vantage isn't the most reliable benchmark in the world? |
|
Lilim
Intel Core i7 920 @4.2GHz HAF 932 - Dual SLI Nvidia GTX 480s 3x Intel X25-M G2 (80GB) SSD RAID0 |
|
rholyrag
Senior Member Joined: 17 Oct 2009 Online Status: Offline Posts: 841 |
Quote Reply Posted: 20 Jan 2011 at 5:31am |
Wow, 141GFlops, and four passes finished in a little over four minutes, very cool. Isn't your 980X only running at 100GFlops?
|
|
"What we do in life echoes in eternity" - Maximus from Gladiator
|
|
Metamorphosis
Newbie Joined: 25 Dec 2010 Online Status: Offline Posts: 62 |
Quote Reply Posted: 20 Jan 2011 at 6:20am |
Actually Dargoonseal I brought it up in another thread.. won't get into it too deep, but of the reviews I can find that shows minimum framerate the 2600k loses out to the 2500k in some game benchmarks. I'm assuming its related to hyperthreading. Though like I said in it .. more testing is needed. I brought it up here because I thought it was worth mentioning since it was about SandyBridge and some quirks its exhibiting.
If you want to ignore it then go for it. I was hoping to get some verification from the community by hopefully getting a larger sample.. not to be completely dismissed and called BS. |
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 20 Jan 2011 at 9:10am |
I will look into it more, but I tried 3 different GPUs, HT on/HT off differences were within the variance that 3Dmark Vantage will have between runs. Then I looked at several different GPU's, with the same clock/RAM speed, but using a wide array of CPU's on HWbot, and no dice same scores within tolerance |
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 20 Jan 2011 at 9:43am |
yep, 100 out of 980x, so for AVX apps, SB will really kick butt, and also makes me pretty hopeful on the next socket, should be a real beast!
|
|
Dragoonseal
DS Veteran Joined: 03 Apr 2009 Online Status: Offline Posts: 2247 |
Quote Reply Posted: 20 Jan 2011 at 11:13am |
Come to think of it, have you tried the 980X with SP1 yet?
|
|
Lilim
Intel Core i7 920 @4.2GHz HAF 932 - Dual SLI Nvidia GTX 480s 3x Intel X25-M G2 (80GB) SSD RAID0 |
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 20 Jan 2011 at 11:22am |
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 20 Jan 2011 at 4:54pm |
Metamorphisis, , if you look up a couple post you will see a blurp about the ultra noobs, using the GPU to calc the CPU portion of 3Dmark Vantage..
Thanks for posting up a couple screenshots of someone like that, I did not know there were even people left that would do that.
|
|
Metamorphosis
Newbie Joined: 25 Dec 2010 Online Status: Offline Posts: 62 |
Quote Reply Posted: 21 Jan 2011 at 12:06am |
... so his CPU is 30degrees higher doing what exactly then?
and there's no reason to call yourself stupid Are you talking about having Physx running? Hes got screenshots up with that disabled too. Otherwise I have no idea what your talking about I've never seen a run CPU tests on GPU as an option. |
|
justin.kerr
DS Veteran Joined: 06 May 2008 Online Status: Offline Posts: 5084 |
Quote Reply Posted: 21 Jan 2011 at 9:05am |
30C higher will lower life a ton, like I just stated above. and depending on a massive amount of variables, effect stability.
Running with PPU on will use the GPU to perform the CPU test. I can explain more if needed.
Vantage between runs will have a variance higher than what he posted the difference between the two screenshots.. so zero meaning there.
|
|
ablahblah
DS Veteran Joined: 12 Jun 2009 Online Status: Offline Posts: 2312 |
Quote Reply Posted: 21 Jan 2011 at 2:19pm |
You know, with what the mainstream SB is bringing to the market, I'm starting to wonder what the heck Intel packed into the high-end line this year
|
|
R4D4RPR00F
Core i7 920 @ 3.9Ghz Asus Sabertooth X58 EVGA GTX 570 Mushkin 6GB 1414Mhz |
|
Post Reply |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |