IntelBurnTest - The new stress-testing program

[AgentGOD]

Don't be afraid, this just makes the long and painful process of running Prime95 much shorter.

[Fungus]

I get access violation when I try to run it, even at stock speeds.



Windows XP SP3 32bit, up to date. DEP=AlwaysOff.

EDIT:

Might be cause of PAE enabling... if I choose level 2 (1/2 mem) it runs fine and passes 10 iterations.

[Brama]

I get access violation when I try to run it, even at stock speeds.



Windows XP SP3 32bit, up to date. DEP=AlwaysOff.

EDIT:

Might be cause of PAE enabling... if I choose level 2 (1/2 mem) it runs fine and passes 10 iterations.

This is a small bug when you use a 32 bit O.S. with more than 2 Gb of memory available and you "TURN ON ERROR DETECTION" and use "Stress level #1"

[Pyr0]

any chance of being able to use flags/switches on a shortcut to run the app quickly without having to answer the questions for error detection etc. please? (no error detect, max stress, 'X' loops)

also, is it possible to list the number of the run (similar to pi) as it would make it a lot easier to quickly see how far it's gotten without having to count the lines when doing 200 loops

TIA

[Justintoxicated]

Anyway we can add time stamps and iteration number? It would be nice to know how long this runs for before becoming unstable! I set mine for 100 runs but since I have 8GB memory it takes hours n hours(vs days and days or never in prime?)...This is why I don't just run 20 interations (takes like an hour with 8GB memory for 20 passes). I left this on overnight, it didn't have any errors for several hours. But I guess I'm still not stable. This program has given a new definition of a stable intel system! You can pass prime for 24 hours....But are you really stable at 4.0ghz on a quadcore?
8GB mememory tresting seems to take exponentialy longer than 4GB testing.

Before I made some Adjustments I was prime stable for 12 hours, IntelBurnTest errored on the 7th pass. Added some voltage to memory, and now this. I would bet that I can pqass prime95 for over 24 hours in this state.

[Dee]

I love this program. Thank you, AgentGOD. Great work.

[BeastNotro]

Justintoxicated that long run is still 2 errors in it so it is still not 100% ultimate stable, but yeah i bet prime95 or Otrthos will be stable in that state 24hrs or more as they are less harsher, but linpack seems to be the best tester for Ultimate Stability for long usage at full load.

In only 5-10 passes and i fold all day with no issues so far and game play too no issues so far.

[Justintoxicated]

Justintoxicated that long run is still 2 errors in it so it is still not 100% ultimate stable, but yeah i bet prime95 or Otrthos will be stable in that state 24hrs or more as they are less harsher, but linpack seems to be the best tester for Ultimate Stability for long usage at full load.
Actualy there were 4 or 5 (but I couldn't fit it all on screen. Another idea would be an error counter, but that is harder to implement heh. If this program was Java based interface and coding i would be glad to help.
In only 5-10 passes and i fold all day with no issues so far and game play too no issues so far.

yea I bumped the memory another .015v this morning running another 100 passes!

[Razmatazz]

I just got through decreasing Vcore until I got an error. I was running 10 iterations. After I got an error I increased Vcore back up 1 notch just to be sure that things remained stable for 10 iterations. I then tried prime95 and get an error within 15 seconds. I tried bumping up Vcore a couple of notches. I am not Prime95 stable and get errors within seconds. Any thoughts/suggestions? Am running Vista64.

[OC_Newbee]

I just got through decreasing Vcore until I got an error. I was running 10 iterations. After I got an error I increased Vcore back up 1 notch just to be sure that things remained stable for 10 iterations. I then tried prime95 and get an error within 15 seconds. I tried bumping up Vcore a couple of notches. I am not Prime95 stable and get errors within seconds. Any thoughts/suggestions? Am running Vista64.

Yeah I was experiencing the same thing when only selecting less than 30 iterations. It would pass IntelBurnTest version 1.4 option 1 (max memory used) on XP x64-edition but error out with prime95 v2.56. Another word is I now make sure that it passes at least 30 iterations with IntelBurnTest v1.4 option 1 and prime95 (@ the very least 12 hours ) stable before considering that my system is good to go.

Without a doubt this program is perfect for short testing your oc (thanks to the aurthor ), but I still use the good ol prime95 for final testing.

[Razmatazz]

Ok I am incredibly bummed. I have as of this point still not truly had a stable overclock ever so it would seem. I have been trying for many hours. I am stable FSB 400 MHz which is default for this processor. Memory is at 1600 MHz. I tried to bump the FSB to 425 MHz and Memory to 1700 MHz. I had thought that I was stable according to BurnTest. I adjusted voltages and found that I needed to add a little VTT and Pll voltage and I also needed to up the NB voltage. I initally pushed up the Vcore a few notches and Burntest ran great. I then decreased the NB voltage until I found where errors occured and went back up one notch on the NB. Burntest was again stable times 10 iterations. I then decreased Vcore until I got an unstable run. I then bumped up the Vcore one notch and again I was Burntest stable times 10 iterations. Now for the frusterating part. I have to question the whole quick test idea. With BurnTest I was stable many times at 10 iterations as I was decreasing Vcore and NB. I would recieve error in Prime95 v25.6 in seconds. I then tried increasing all of the voltages that I had just got done decreasing. I increased Vcore, NB, VTT, Pll and each time I would get errors in Prime95 within 10 seconds on multiple cores. I then put the system back to FSB 400 MHz and now both BurnTest and Prime95 both run without errors. I would love to be able to overclock this rig but man, this has been a real headache. I would be most grateful for assistance.

[Loser777]

There isn't an end all test... if even one test fails, it means your system isn't stable.
But, don't complain that this is an issue with a specific stress test... post you own thread if you need help with an overclock.

[Justintoxicated]

Only thing I can say is I don't think prime95 tests the memory as much as burntest. Or the CPU really, but I don't know what burntest is doing at all (magical). But thats my experience and aparently others vary. Make sure you have another program open like realtemp so can can make sure your CPU is not thermal throtteling or varying in speed while running these stress tests.

[martymonster]

After running (and passing with NO errors) Prime95 for 25 hours, then 400 passes of Intelburntest V1.4 ( 10.5 hrs) and 5 runs of 3DMark06 my system was still not stable in normal use.
My PC had been on for nearly 50 hours, a few hours after both those tests I went to open something and my PC froze.
See picture.
I have sinced upp'd the NB, PLL, FSB Term voltages and will only be satisfied when in normal use my system does not hang.
The above programs are great for getting the VCore correct but I find that nothing compares to actually using the system in daily use.
Previously when my FSB was 380 (now 345) and my NB etc Volts were a lot higher, this type of system freeze could occur when opening a window, opening a program or even when shutting down the system.

[AgentGOD]

In that case, have you tried a clean install of Windows? It sure sounds like software issue, not hardware. Also, did you try Memtest86+ to test for memory issue(s), and ATITool to test for GPU artifacting?

[billdavis]

After running (and passing with NO errors) Prime95 for 25 hours, then 400 passes of Intelburntest V1.4 ( 10.5 hrs) and 5 runs of 3DMark06 my system was still not stable in normal use.
My PC had been on for nearly 50 hours, a few hours after both those tests I went to open something and my PC froze.
See picture.
I have sinced upp'd the NB, PLL, FSB Term voltages and will only be satisfied when in normal use my system does not hang.
The above programs are great for getting the VCore correct but I find that nothing compares to actually using the system in daily use.
Previously when my FSB was 380 (now 345) and my NB etc Volts were a lot higher, this type of system freeze could occur when opening a window, opening a program or even when shutting down the system.

and what os are you using

[billdavis]

some work

[BAR*B*Q]

i testet my system with 8 gb with burntest several times. its allways stable but with prim95 i get a crash with a reboot.

[Razmatazz]

There isn't an end all test... if even one test fails, it means your system isn't stable.
But, don't complain that this is an issue with a specific stress test... post you own thread if you need help with an overclock.

You my friend have missed the point. This is exactly the place to post our experiences with this program and how it compares to other tests, especially if the program is behaving differently then expected. I'm sure the author of this program and others would like to know even if you don't.

That's the old Prime95 rule. Refer to my experience in my first post in this thread. 10 hrs Prime95 stable = failed Linpack in 2 minutes.

So it would seem that BurnTest should give an error sooner than other stess testing programs. I only increased a small amount from default 3.2 GHz to 3.4 GHZ and was BurnTest stable (multiple runs of 10 iterations). I, however, get Prime95 errors in multiple cores within seconds at this speed. Any thoughts here?

[AgentGOD]

Are you using 32-bit or 64-bit Windows?

i testet my system with 8 gb with burntest several times. its allways stable but with prim95 i get a crash with a reboot.

Have you tried using the "Custom" option, and entering 3100 MB for testing?

[Justintoxicated]

billdavis, very nice, I wish I could get my Q9450 stable at 3400MHZ 425FSB...I can't pass 30 interations of Burntest with 8GB memory, with 4Gb it passes fine though. Really struggeling with this I need to get this PC working I have alot of other thigns to do. It's driving me crazy, I seem to be prime stable I just can't pass teh burn test.

[WaterFlex]

AgentGOD
Thank you very much ! U make my life faster and more interesting

[AgentGOD]

Updated to v1.5
- Updated to latest Linpack binaries
- Updated credits section of readme
- Fixed a crashing bug when using new binaries
Note: This version will not accept older binaries, and older versions will not accept the
new binaries. Each situation will cause the program to crash.

Note: I can't tell you how reliable the new binaries are, since they are untested. Please do test

I've also updated the native 64-bit version for WinPE.

[martymonster]

and what os are you using

As stated in my previous post,
Vista 64
Q6600 G0
PSU - Corsair HX1000
DDR2 - OCZ 2 x 2GB 6400 @ 5-4-4-15 @1.9V
(Was 4 x 2GB but Memtest found 1 bad stick so 2 have been returned and waiting for replacement)

System is water cooled in a Coolermaster Cosmos S case.
GPU is an ASUS 3870 which is NOT overclocked.

In a months time I may get some free time to reinstall the OS and do some testing before restoring all my apps.

[003]

Updated to v1.5
- Updated to latest Linpack binaries
- Updated credits section of readme
- Fixed a crashing bug when using new binaries
Note: This version will not accept older binaries, and older versions will not accept the
new binaries. Each situation will cause the program to crash.

Note: I can't tell you how reliable the new binaries are, since they are untested. Please do test

I've also updated the native 64-bit version for WinPE.

AgentGOD, may I ask where you can find the latest linpack binaries? I have been checking this link:
http://www.intel.com/cd/software/pro...eng/363184.htm

But I see it is the same it has always been. Where can you always find the most up to date binaries? I searched the intel developer site but was unable to find them.

EDIT:
It would be nice if you could keep a running changelog in your OP