IntelBurnTest - The new stress-testing program

[BeastNotro]

This could be my PC being unstable....but if I run 7 tests using 3300MB of RAM (as I am on Vista x64 SP1) all values are identical (which is good as that indicates stability), however if I run, then a few seconds later open CPU-Z and RealTemp while the test is in progress my 1st 2 results are completely different. Is this a sign of instability? Or is it because the applications I am launching whilst the test is in progress are causing some sort of conflict.
Please note that if I run Real Temp and CPU-Z and leave them open I can complete 7 tests with no errors, however launching them during the test causes 2 results to differ.
John

If in all tests the numbers are identical then your OK. meaning

EXAMPLE ONLY:

1.20 = 3.23
1.20 = 3.23
1.20 = 3.23
1.20 = 3.23
1.20 = 3.23

Next time you have more free memory to test:

1.27 = 3.76
1.27 = 3.76
1.27 = 3.76
1.27 = 3.76
1.27 = 3.76

Basically the test is based upon FREE memory (stress CPU stability) and if you open a program up while it is running then that would lower the memory being used for the test.

Open all programs you want open first then check FREE memory you have left == open task manager to see what is FREE and then use that number in the linkpack test.

As long as you have the numbers identical NO two test maybe the same, but as long as they are identical numbers.

You can't have say ....

EXAMPLE ONLY:

1.20 = 3.23
1.10 = 3.85
1.20 = 3.22
1.15 = 3.23

As that would be unstable.

Here is my test see the numbers are identical in both rows ( THE RESIDUAL ROWS) so i pass and i ran it again used more free memory different numbers again .... but identical in both rows:

http://www.xtremesystems.org/forums/...7&postcount=81

Now look at Super Nades .... his RESIDUAL ROWS not identical:

http://www.xtremesystems.org/forums/...7&postcount=80

[Super Nade]

Hmm..I've never had a single BSOD or hint of instability thus far. But, it is always good to know that I need to work a bit more on my rig. I know this is XS, but I leave the BSOD loops to my secondary benching windows install...lol

[BeastNotro]

Hmm..I've never had a single BSOD or hint of instability thus far. But, it is always good to know that I need to work a bit more on my rig. I know this is XS, but I leave the BSOD loops to my secondary benching windows install...lol

LMAO.. good to have another OS HD for corruption only... lol

But Linpack is a good stability 100% data stability tester in which at anytime that if your not stable .. you could/will get data corruption ... maybe not this week, maybe in 2 weeks ... maybe a month .. maybe 2 months.... and a format will be needed.

I like stability and if this program can tell me if i am and not that awesome.. also don't stop using Orthos either ..... use both Orthos and This from now on.. Intel CPU's only ... LOL

[Xello]

So how important are these cases where people are prime stable but not stable in this test?

I thought that if you were prime stable you would essentially not experience any issues that were down to instability in your system (talking about general home pc use here), how important is this test for general users?

[BeastNotro]

I think its about how stable do you really want to be is all and the more the merrier is how i look at it.

[Pyr0]

mmm toasty

[RealTelstar]

Program Creator, If you created a program that would automatically run 2 tests, 1 with low ram, and one with ~remainder free ram, and also check for errors, that would be donation worthy. As the program stands right now, I agree with krille.

This is exactly what i'm looking for. Better interface all-in-one program.

[BeastNotro]

Pyr0

Awesome looks to be stable as results are Perfect.

Now 71C is that all you can do.....

I need to get lower temps on full load....

[WoZZeR999]

For general users, it's about as important as being 48 hour prime stable, both small and blend. In other words, it's another fun test to run.

[Loser777]

I don't get the point of this program... anyone with half a brain and a DOS GUI doesn't need this...
Use Far Manager...

[BeastNotro]

I don't get the point of this program... anyone with half a brain and a DOS GUI doesn't need this...
Use Far Manager...

I like it as it is simple, fast and useful....

[Loser777]

Yeah, the method that I was talking about is just as fast...

[Leeghoofd]

I never used linpack as I have never made time to learn it, let alone trying to configure it... I'm not intereseted how it works or how it is build up... we got other dudes like you that er interested in those things... I just test it and kinda like this nifty program for a 10 minute test...and I think many will follow my point of view...

I'm a hardware builder and love to OC (max on water), for the rest it's gaming gaming and gaming...

[Loser777]

For it to be a real program I think it would need to have more features, rather than just be a program that displays an output txt file... come on...

[WoZZeR999]

Yes, you like far manager, we know. It's a front end for Linpack. That's all it was claimed to be. We're trying to help the program creator out so that he can make it into a full program.

[btdvox]

I usually test Prime stable for 6 hrs on small ffts and then blend, but I actually found that 2 hour tests on each sect. of OCCT found errors for me and use that from now, and neverh ad a problem, Plus its less time consuming.

Though I never leave my computer on 24/7 but for games I never game for more than 2 hrs at a time anyways so...

[AgentGOD]

Well thank you guys for your suggestions (except a certain someone's [*pointing at Loser777*]).

You should at least give Cronos proper credit in the first post.

Well, I gave him the well deserved credit for the equation for calculating the size to use for specified memory, just look at the header of the program. My apologies for neglecting to point it out in the readme and the first post.

[HDCHOPPER]

thanks Agent GOD for shareing (free) a good program and trying to better it also
(the best thing 1 can do for a bunch of overclockers )

but can it be had in a plain ole zip file?

[AgentGOD]

thanks Agent GOD for shareing (free) a good program and trying to better it also
(the best thing 1 can do for a bunch of overclockers )

but can it be had in a plain ole zip file?

You're welcome. Sure, I'll make the next release be in a .zip format.

...automatically run 2 tests, 1 with low ram, and one with ~remainder free ram.

Instead of doing that, I have it allow you to select the "stress level" of the test you wish to run, which varies by RAM allocation amount based on available memory. The new version can already automatically determine the available physical memory.

I did not follow your suggestion to run two tests as said above, because Linpack is the most accurate when using the most memory possible. It brings maximum CPU stress as well as maximum RAM stress, which also means the most accuracy. I added the option to select the amount of stress as some people want to do quicker tests rather than 8 minute ones.

[WoZZeR999]

After some testing, I realized that running with a small amount of ram isn't productive like prime. Prime with Small FTT's is good for testing just CPU stability, while Blend/Linpack are better for complete subsystem stability. Both deserve their own place as they test different things. 'IntelBurnTest' just needs the 1 max ram test. Error checking and % completed successfully would be helpful too. If after running it I see 90% correct, that may be fine, but if I see 30% correct then that would be trouble.

[BeastNotro]

I always run Orthos Blend no matter what on any CPU i have OCING or even stock if need to test stock for at least 8-10 hours and at times 12-16 hours if i feel like it... lol

For Linpack i used 1857, 2000 and 2200 memory for testing and all passed.

Even this overclock on my q6600 as seen on page 4 (Linapack test) of this thread i used Orthos Blend, Linkpack says it is at 4050mhz.... lol when it is only at 3.6ghz.

If you get an error of SUM in Orthos Blend it usually means CPU related, but at times it can be NB volt related or PLL.

If it is a rounding ERROR then usually it is memory related whether it is the CPU controller flaking out or memory timings or voltage ..... or even NB voltage.

8/10 SUM error is CPU related.
8/10 Rounding error is Memory related.

[AgentGOD]

Even this overclock on my q6600 as seen on page 4 (Linapack test) of this thread i used Orthos Blend, Linkpack says it is at 4050mhz.... lol when it is only at 3.6ghz.

Both Linpack and Windows assumes default multiplier, thus showing higher speed than what you actually have it set at.

Also, Linpack is much more accurate, way beyond Prime95 when testing under 64-bit Windows. Read my personal experience in the first post to get a glimpse of what I'm talking about.

After some testing, I realized that running with a small amount of ram isn't productive like prime. Prime with Small FTT's is good for testing just CPU stability, while Blend/Linpack are better for complete subsystem stability. Both deserve their own place as they test different things. 'IntelBurnTest' just needs the 1 max ram test. Error checking and % completed successfully would be helpful too. If after running it I see 90% correct, that may be fine, but if I see 30% correct then that would be trouble.

As I said before, please read my personal experience in the first post.

Summary of it (using the current rig I have):
- Prime95 Small FFTs and Blend stable at 3.5 GHz, 1.42v for at least 15 hours, while Linpack, using most of my memory, started BSODing.
- Raised vcore higher (to 1.45v), no more BSOD, but Residual(norm) values weren't all equal.
- Raised vcore higher yet again to 1.475v, One out of 10 Residual(norm) values was inconsistent
- Raised vcore to 1.496v, all Residual(norm) values equal, even after running 10-tests.

This was all done on XP X64 SP2.

[AndrewL1950]

My room is extremely hot right now, about 32C. I ran this program with 100mb dedicated, it all the residuals matched after 5 passes. Then I ran it with 2000mb of ram dedicated, and my temps went up to about 79C. ONe thing is clear, this small Noctua needs to be replaced, ASAP. I am at stock volts also, jesus.

[AgentGOD]

My room is extremely hot right now, about 32C. I ran this program with 100mb dedicated, it all the residuals matched after 5 passes. Then I ran it with 2000mb of ram dedicated, and my temps went up to about 79C. ONe thing is clear, this small Noctua needs to be replaced, ASAP. I am at stock volts also, jesus.

Don't worry too much, I'm getting 89*C on two cores, and 81*C on other two on my Q6600 running Linpack w/ most memory available.

P.S.: I'll be releasing version 1.2 soon when I get back. It's already finished. Keep watching

[BeastNotro]

My room is extremely hot right now, about 32C. I ran this program with 100mb dedicated, it all the residuals matched after 5 passes. Then I ran it with 2000mb of ram dedicated, and my temps went up to about 79C. ONe thing is clear, this small Noctua needs to be replaced, ASAP. I am at stock volts also, jesus.

Read my posts:

http://www.xtremesystems.org/forums/...7&postcount=81

http://www.xtremesystems.org/forums/...&postcount=101