IntelBurnTest - The new stress-testing program

Printable View

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

08-24-2008, 09:41 PM
MaxT

it tells me there is not enough memory when i try to select the max memory test, weird.
08-24-2008, 10:11 PM
AgentGOD

Quote:

Originally Posted by MaxT

it tells me there is not enough memory when i try to select the max memory test, weird.

That happens because you have more than 2 GB of memory and you are using 32-bit Windows. To get around this, use less memory to test, or upgrade to XP 64-bit.
08-24-2008, 10:53 PM
003

I can second XP x64... it is a great OS. I am only using Vista x64 because I have 4 GPUs :(
08-25-2008, 12:57 AM
pest

Quote:

Originally Posted by AgentGOD

That happens because you have more than 2 GB of memory and you are using 32-bit Windows. To get around this, use less memory to test, or upgrade to XP 64-bit.

In one of the previous versions (1.3) i was able to test 2GB on my 32-Bit WinXP.
With 1.5 i'm not able to test more than 900mb

anyway thanks for the little proggie :)
08-25-2008, 01:16 AM
WaterFlex

AgentGOD
Would u like to make the new interface and design for 1.5.1 ? :) I love 1.5.1 ;)
08-25-2008, 06:37 AM
SPDIF_IO

Thanks agent will make a donation tonight when i get home

http://i243.photobucket.com/albums/f...apudding-1.jpg
08-25-2008, 11:32 AM
ahmad

Regardless, if I could do the same test in HCI and have it produce errors FASTER than this tool, why would I use this tool to begin with is my point.

I want a tool that produces errors the fastest. And for memory, this doesn't do it.

HCI - 1024MB errors out FASTER than this tool with 10 passes at 1500MB.

So like I said, its possible that this tool can detected CPU errors FASTER than Prime or SuperPI, but for RAM errors it needs MORE TIME.

:)

Quote:

Originally Posted by BeastNotro

If you not choose the error read out then you will see the output of the testing being done... the error out put is just to tell NEWBIES if they passed or not.

Its not a memory Error testing procedure only error of pass or fail.

Now as for it being CPU stability only = NO as it will test memory as well and it is always suggested and recommened to use ALL available memory to run the test and run it 10 loops (3-6) for a more aggressive testing.

AgentGod put something like i just posted within the testing TEXT so ppl don't get all confused... also actually take out the min and half memory options.

Just leave MAX or Custom .. thats all that is needed.
08-25-2008, 12:01 PM
Warboy

I want a FLOPs printout like the old version had.
08-25-2008, 12:25 PM
BeastNotro

Quote:

Originally Posted by [cTx]Warboy

I want a FLOPs printout like the old version had.

Just don't check off error reporting as all that is for is to tell you if you passed or not.
08-25-2008, 02:53 PM
Razmatazz

Any ideas why I would be stable with BurnTest and get errors within seconds with Prime95? I am stable using both at default 3.2 GHz but get errors at 3.4 GHz with Prime95 even though BurnTest remains stable.
08-25-2008, 02:56 PM
AgentGOD

Quote:

Originally Posted by Razmatazz

Any ideas why I would be stable with BurnTest and get errors within seconds with Prime95? I am stable using both at default 3.2 GHz but get errors at 3.4 GHz with Prime95 even though BurnTest remains stable.

Use 64-bit OS (XP x64 SP2 preferred), test with maximum memory, and do 20 iterations.
08-25-2008, 04:02 PM
MaxT

Quote:

Originally Posted by AgentGOD

That happens because you have more than 2 GB of memory and you are using 32-bit Windows. To get around this, use less memory to test, or upgrade to XP 64-bit.

Got it. Thanks!
08-25-2008, 06:08 PM
WaterFlex

AgentGOD, do u see my message above?
08-26-2008, 04:15 AM
Razmatazz

Quote:

Originally Posted by AgentGOD

Use 64-bit OS (XP x64 SP2 preferred), test with maximum memory, and do 20 iterations.

I am running 64 bit Vista. I have not run 20 iterations at one time but have run 10 iterations multiple times after minor changes and have been stable with BurnTest but still run into early error (less than 15 seconds) with Prime95 with only a minor overclock (3.2 GHz to 3.4 GHz). I had thought Burntest would find instability faster then other stress testing programs. Is this a 64-bit Vista issue? Any other thoughts?
08-26-2008, 08:59 AM
AgentGOD

Quote:

Originally Posted by Razmatazz

I am running 64 bit Vista. I have not run 20 iterations at one time but have run 10 iterations multiple times after minor changes and have been stable with BurnTest but still run into early error (less than 15 seconds) with Prime95 with only a minor overclock (3.2 GHz to 3.4 GHz). I had thought Burntest would find instability faster then other stress testing programs. Is this a 64-bit Vista issue? Any other thoughts?

Do you have voltage variations running IntelBurnTest vs running Prime95? (Talking about CPU voltage at load)

If so, that would be what I'd look into, especially with Load-Line Calibration on. It seems that some boards give the CPU different voltages during different loads. I have not tested Vista 64-bit on this PC, so I couldn't tell you if that's the reason. I've only tested XP x64 Edition SP2 (so I can't tell you how effective it is on Vista 64-bit), and from my experience, it's much much faster than Prime95. My original 10hr Prime95 stable CPU clock failed IntelBurnTest within 2 minutes. For that test, I used the 64-bit version of Prime95.

Quote:

Originally Posted by WaterFlex

AgentGOD, do u see my message above?

Yeah, I saw your message about slapping a full Windows GUI on it vs console, but I decided to stick with the original style.
08-26-2008, 09:51 AM
WaterFlex

Quote:

Originally Posted by AgentGOD

Yeah, I saw your message about slapping a full Windows GUI on it vs console, but I decided to stick with the original style.

Good luck then! I`m fond of this test, i love to burn my PC :D.
I`m looking forward to the new versions ;)
08-26-2008, 10:07 AM
Justintoxicated

Quote:

Originally Posted by Razmatazz

I am running 64 bit Vista. I have not run 20 iterations at one time but have run 10 iterations multiple times after minor changes and have been stable with BurnTest but still run into early error (less than 15 seconds) with Prime95 with only a minor overclock (3.2 GHz to 3.4 GHz). I had thought Burntest would find instability faster then other stress testing programs. Is this a 64-bit Vista issue? Any other thoughts?

Very odd! I have no idea! I hope someone can chime in!
08-26-2008, 01:24 PM
KoHaN69

http://img403.imageshack.us/img403/6008/40710998ev9.jpg

http://img403.imageshack.us/img403/9948/44180661er9.jpg

:shrug:
08-26-2008, 02:13 PM
AgentGOD

"Weird events" can be caused by system instability as well. I learned my lesson with my AMD Opteron... it wasn't fully stable, and my Windows installations all eventually got screwy like yours (both XP and Vista [I had both installed in a dual-boot configuration, using 32-bit]).

Keep this in mind: neither IntelBurnTest nor Intel(R) LINPACK sets anything to start on Windows startup, nor start stress testing without the user's consent.
08-26-2008, 07:15 PM
squito

Quote:

Originally Posted by Razmatazz

Any ideas why I would be stable with BurnTest and get errors within seconds with Prime95? I am stable using both at default 3.2 GHz but get errors at 3.4 GHz with Prime95 even though BurnTest remains stable.

I'm good on both BurnTest and Prime95 at 3.6 GHz - no go on SuperPi 32M at any clock, stock or otherwise - 64 bit Vista here also :rolleyes:
08-27-2008, 03:13 AM
AgentGOD

Can you pass Memtest86+ for a few hours?
08-27-2008, 08:31 AM
ahmad

Quote:

Originally Posted by squito

I'm good on both BurnTest and Prime95 at 3.6 GHz - no go on SuperPi 32M at any clock, stock or otherwise - 64 bit Vista here also :rolleyes:

run this: http://hcidesign.com/memtest/

use two instances, at 1024MB each (if you have more than 3GB of RAM), and do at least 300% coverage.
08-27-2008, 08:51 AM
RejZoR

Quote:

Originally Posted by squito

I'm good on both BurnTest and Prime95 at 3.6 GHz - no go on SuperPi 32M at any clock, stock or otherwise - 64 bit Vista here also :rolleyes:

Thats because SuperPi isn't compatible with Vista (at all). Use HyperPi instead...
Same thing as SuperPi with some extra Vista compatibility fetaures.
08-27-2008, 09:41 AM
Razmatazz

Quote:

Originally Posted by AgentGOD

Do you have voltage variations running IntelBurnTest vs running Prime95? (Talking about CPU voltage at load)

If so, that would be what I'd look into, especially with Load-Line Calibration on. It seems that some boards give the CPU different voltages during different loads. I have not tested Vista 64-bit on this PC, so I couldn't tell you if that's the reason. I've only tested XP x64 Edition SP2 (so I can't tell you how effective it is on Vista 64-bit), and from my experience, it's much much faster than Prime95. My original 10hr Prime95 stable CPU clock failed IntelBurnTest within 2 minutes. For that test, I used the 64-bit version of Prime95.

Voltages are exactly the same when running BurnTest and Prime95. I'm still at a loss with this one. Perhaps it is a Vista issue?

Quote:

Originally Posted by squito

I'm good on both BurnTest and Prime95 at 3.6 GHz - no go on SuperPi 32M at any clock, stock or otherwise - 64 bit Vista here also

Quote:

Originally Posted by RejZoR

Thats because SuperPi isn't compatible with Vista (at all). Use HyperPi instead...
Same thing as SuperPi with some extra Vista compatibility fetaures.

Thanks for the info on SuperPi vs HyperPi. I found that I was 3D unstable and SuperPi unstable until I adjusted clock skews. I am currently at a delay of 250ps for both channels. I am now SuperPi and 3D stable with 4GB memory at 1600 MHz.
08-27-2008, 12:44 PM
squito

Quote:

Originally Posted by AgentGOD

Can you pass Memtest86+ for a few hours?

I ran the through the whole test twice for what it's worth - seemed like it took couple of hours to yours truly :rolleyes:

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