Official Intel Gulftown (Westmere) Overclocking and Batch Tracking Thread

Printable View

Show 100 post(s) from this thread on one page

03-21-2010, 02:33 PM
lawtq

Quote:

Originally Posted by SteveRo

I thought I might post this in the intel forum also - good voltage data - plus another opportuity to indulge vanity.

Ok, so new pcmark vantage WR - I had to push to 1.45 vcore but I wanted to be the first over 30k -

http://img408.imageshack.us/img408/3...agewr30359.png

Please post a pic of the inside of you PC, especially the 2x boreas
setup!! Loving the temps!
03-21-2010, 02:40 PM
Mech0z

Quote:

Originally Posted by Musho

Because he's using 6 cores, plus a kickass storage system. PCMark Vantage is _highly_ dependent on a fast storage system.

Oh my bad, thought it was 3dmark vantage :fight:
03-21-2010, 03:12 PM
SteveRo

Quote:

Originally Posted by Mech0z

How the hell can you beat Vantage record without ln2 :o aka "only" 4600mhz compared to the 6500 others have hit

Also, a good SS overclock should have no problem doing 5+ghz with a 980x.
03-21-2010, 03:56 PM
Rubycon

Quote:

Originally Posted by Sam_oslo

I don't know what's possibly wrong with LinX, but it seams not be using all 6C/12T. In case, it won't stress 100% and won't be able to determine the 24/7 stability either. Running 8 hours Prime SamllFFT (or at least a couple of hours) would be much better indication then. At least until they find what's wrong and fix the LinX.

LinX is messing up on dual Xeon systems too. I thought I had a throttling issue when I saw the GFLOPS go down.

It seems like it will go up to 2048MB size. After that it's a crap shoot. If you see your temps not as high or power draw lower than you know it's not stressing the CPU(s) despite all programs, task manager, etc. showing 100% load.

OCCT still works, however. Faster for finding errors than Prime95.

Until Linpack is updated to support these new CPUs we're stuck with smaller problem size (keep the RAM used to 2GB or lower which is not as thorough!) or use OCCT.
03-21-2010, 06:29 PM
[200]

Have you considered keeping an eye on the Extreme Overclocking section of the forum? Maybe you could add liquid nitrogen to the list?
03-21-2010, 11:21 PM
zalbard

Quote:

Originally Posted by tool_462

I'd tell you the batch on the chip but its under my block and I don't think anyone cares about ES batches :p:

Sure we do! :up:

Quote:

Originally Posted by Sam_oslo

I don't know what's possibly wrong with LinX, but it seams not be using all 6C/12T. In case, it won't stress 100% and won't be able to determine the 24/7 stability either. Running 8 hours Prime SamllFFT (or at least a couple of hours) would be much better indication then. At least until they find what's wrong and fix the LinX.

I'll give an option of 2 hours of Small Ffts for now.

Quote:

Originally Posted by SteveRo

What I'm getting at is it should hopefully still be a measure of stability -perhaps just not as good as for other procs?

Yeah, that's how I see it as well.

Quote:

Originally Posted by hlonipha

A little higher on the volts but will start to decrease and see if I can get it stable with a bit lower.
Image

Added yours! :up:

Quote:

Originally Posted by [200]

Have you considered keeping an eye on the Extreme Overclocking section of the forum? Maybe you could add liquid nitrogen to the list?

I'll try! :up:

Regarding LN2... Feel free to point me in direction of new submissions in various threads in case I miss those, will make my job much easier. Thank you! :up:
03-22-2010, 06:34 AM
SteveRo

Good morning Mr Zalbard,

I see you added a 3dMark vantage catagory - shouldn't we also have a pcmark catagory?
03-22-2010, 06:41 AM
zalbard

Quote:

Originally Posted by SteveRo

Good morning Mr Zalbard,

I see you added a 3dMark vantage catagory - shouldn't we also have a pcmark catagory?

We could, I guess... Do you think it will be useful?
I have nothing against PCMark.
I just added 3 most popular 'benchmarks' on LN2. I couldn't come up with one every single bencher uses (and didn't want to use like 20 - too much pointless work), but most either bech 2D (surely SPi 1M in this case) or 3D (3DM Vantage usually), and some also come up with max validation.
Not many play with PCMark Vantage for some reason (expensive storage systems I guess).
I think both Vantage benchmarks cause similar CPU stress levels, with 3DMark being more popular. So I opted for that.
03-22-2010, 02:24 PM
SteveRo

Quote:

Originally Posted by zalbard

We could, I guess... Do you think it will be useful?
I have nothing against PCMark.
I just added 3 most popular 'benchmarks' on LN2. I couldn't come up with one every single bencher uses (and didn't want to use like 20 - too much pointless work), but most either bech 2D (surely SPi 1M in this case) or 3D (3DM Vantage usually), and some also come up with max validation.
Not many play with PCMark Vantage for some reason (expensive storage systems I guess).
I think both Vantage benchmarks cause similar CPU stress levels, with 3DMark being more popular. So I opted for that.

Your right - not that many use pcmark vantage, 3dmark is probably fine.
I do think pcmark vantage is gaining in popularity - its a better measure of all around pc performance - but the fact that you have to pay for it is holding in back, same for 3dmark vantage but perhaps to a lessor extent. Also it takes longer to run pcmark - people don't like to wait.
03-22-2010, 02:26 PM
SteveRo

Quote:

Originally Posted by lawtq

Please post a pic of the inside of you PC, especially the 2x boreas
setup!! Loving the temps!

Good evening, not that much to look at really -

see - http://www.xtremesystems.org/forums/...51&postcount=6
03-22-2010, 04:55 PM
railmeat

[IMG]http://i76.photobucket.com/albums/j1...oto-0029-1.jpg[/IMG]

http://i76.photobucket.com/albums/j1...owsscore-1.png
http://i76.photobucket.com/albums/j1...owsscore-1.png

my win7 score ran last week with my 920...my 980 only got 1 point higher?hmmm

my c300 screams tho --128 gig crucial
[IMG]http://i76.photobucket.com/albums/j1...hwin7score.png[/IMG]
03-23-2010, 08:16 AM
Rubycon

LinX 0.64 WILL stress all cores up to 512MB test size. Anything bigger and it generates less heat and GFLOPS.

I've tested this with two different boards and 980Xs.
03-23-2010, 05:11 PM
managerman

Quote:

Originally Posted by Rubycon

LinX 0.64 WILL stress all cores up to 512MB test size. Anything bigger and it generates less heat and GFLOPS.

I've tested this with two different boards and 980Xs.

You mean like this?? :)

Gflops way up..and heat way up...:yepp:

http://i139.photobucket.com/albums/q...12gulftown.jpg
03-23-2010, 08:05 PM
Rubycon

Right...now change the testing from number of tests to run to minutes and move it to 300. That will run it for five hours. If it does not err out after five hours it probably won't after 20. ;)

[QUOTE=managerman;4302817]You mean like this?? :)

Gflops way up..and heat way up...:yepp:
03-24-2010, 12:06 PM
MaximumPig

got my 980x 3005F215 (from Fry's in the US, ordered 3/22), putting it in my GA-X58 UD7 tonight, will report back. im just a humble (but brave) air-cooler though.
03-24-2010, 08:01 PM
trn

Great thread Zalbard! :up: Might I suggest adding XS user names in post #2 along with the other data? I like to read the info by users, for a lot of people that post here a lot you already know what kind of setup they are running. Anyways... Thought Id make a suggestion before the thread got to long to add or modify the data.

Also, I'm seeing Overclocks without batch #'s and information about the cooling???:slap: This thread is about collecting data about Gulftown/Westmere Batch # overclocking results.
03-25-2010, 01:29 AM
zalbard

Quote:

Originally Posted by trn

Great thread Zalbard! :up: Might I suggest adding XS user names in post #2 along with the other data? I like to read the info by users, for a lot of people that post here a lot you already know what kind of setup they are running. Anyways... Thought Id make a suggestion before the thread got to long to add or modify the data.

I can, I just don't think it's important since it's just about finding out which batch is better, if you want some competition you can probably check my LinX club. :p:

Quote:

Originally Posted by trn

Also, I'm seeing Overclocks without batch #'s and information about the cooling???:slap: This thread is about collecting data about Gulftown/Westmere Batch # overclocking results.

It's not a mandatory thing. It's a general Gulfy OC thread, batch tracking is the secondary goal.
However, I encourage you guys to list those so I can populate the table. :)
03-25-2010, 03:54 AM
SteveRo

Mr Zalbard, Please update the first post, the same chip - 3003B287 - thanks!

http://img682.imageshack.us/img682/9...linxstable.png
03-25-2010, 04:06 AM
zalbard

Quote:

Originally Posted by SteveRo

Mr Zalbard, Please update the first post, the same chip - 3003B287 - thanks!

Done! :up:
03-25-2010, 04:32 AM
ci2kla

Quote:

Originally Posted by SteveRo

Mr Zalbard, Please update the first post, the same chip - 3003B287 - thanks!

http://img682.imageshack.us/img682/9...linxstable.png

What a killer batch!:eek:
03-25-2010, 04:45 AM
Sam_oslo

Quote:

Originally Posted by ci2kla

What a killer batch!:eek:

Yeah, really good result. It looks like the return of B-batch.
03-25-2010, 04:49 AM
pinto

Don't forget that he didn't use HT and use some really cold water.
03-25-2010, 04:52 AM
SteveRo

Thanks guys, I'm thinking of trying 1.5v but I really don't want to make an entry in the 32nm graveyard thread.
03-25-2010, 04:56 AM
SteveRo

Quote:

Originally Posted by pinto

Don't forget that he didn't use HT and use some really cold water.

Initial indication is HT on for 32nm doesn't cause as much heat as it did for 45.

Also - droping from ~1.65 down to 1.45vcore, that combined with the 32nm part already runing cooler vs. the 45 - the 2xboreas is working out great.
A little antifreeze keeps the water from freezing.
When summer comes I will need to be carefull about condensation - I haven't really protected anything in my current setup.
03-25-2010, 06:18 AM
tool_462

Quote:

Originally Posted by SteveRo

Initial indication is HT on for 32nm doesn't cause as much heat as it did for 45.

But it does make the chip not clock as high when enabled, which I think was his point.

Show 100 post(s) from this thread on one page