XStress - A CPU burn-in/stress testing tool

[AgentGOD]

XStress Readme file
--------------------
Version 1.0
Homepage: http://www.ultimate-filez.com
Type: freeware
Author: AgentGOD

How it works:
It stresses your PC's CPU to maximum by having it calculate a very high factorial. This has proven to be more stressful than Prime95.
Remember, monitor your system temperatures wisely, and don't let your CPU fry!

Requirement(s):
Microsoft® Windows 98SE/ME/2000/XP/2003/Vista/XP x64
AMD Athlon(tm) Class CPU (or higher) / Intel® Pentium(tm) Class CPU (or higher)
64 MB System RAM
Microsoft .NET Framework 1.1 (or higher)

Standard Usage:
Type in the amount of threads to use, and type in the digits (in trillions).
Ex: 2, 14 would give you 2 threads, and 14 trillion factorial calcuations.

This program was based on Prime2006 (an older version of the stress testing program).


http://www.ultimate-filez.com/xstress.jpg

Download Link:
Here (http://www.ultimate-filez.com/XStress.zip)

[AgentGOD]

Small Update:
Shows time elapsed after calculating (cancel or completed)

[DAK1640]

nice find

[AgentGOD]

Not to offend you or anything, but I think you missed out the author name...
:stick:

Another update!
Now supports any amount of threads (as long as you have sufficient system memory, and your CPU can take the heat)

[WeStSiDePLaYa]

i like the multi thread.

but im skeptic of wether it can stress more than SNM.

[AgentGOD]

Before burn-in: Stable at 2721 MHz @ 1.57v

After burn-in: Stable at 2732 MHz @ 1.55v
:banana:

[[cTx]Philosophy]

how long of a burn have you used, and im sure this is gonna come up, so ill go ahead and throw it in first

A64 cpu's cannot be burned in, whats actually happening is that the thermal compound between the IHS and the core takes at or around 200 hours to get up to operating thickness to be able to make good enough contact with the core to make someone think they've actually burned in their cpu to a higher lever when actually all youve done is break in the thermal compount..

Im in no means downing your proggie or ur overclock in anyway dude, I have actually downloaded it myself, I just knew this would eventually come from someone here because this has been a hot topic here for a long time, I just wanted to beat someone else to the punch :)

[masterofpuppets]

Will this put as much stress on the system as dual Prime95 at priority 9 with 1024-4096 in-place FFT's? Not gonna switch until I know, since that's my primary method of stability testing right now.

EDIT: Still gonna try it out anyway though :)

[pete5990]

I don't like Prime95. I've managed to run it for 16 hours without an error, but then ran Aquamark 3 and my PC crashed just 30 seconds into it.

[lawrywild]

I don't like Prime95. I've managed to run it for 16 hours without an error, but then ran Aquamark 3 and my PC crashed just 30 seconds into it.

maybe psu couldnt handle the extra load your gfx card was putting on it

[AgentGOD]

To prove this ain't usless:

http://www.ultimate-filez.com/images/cpu_burn.jpg
:stick:
(Higher the better, in this case, the purpose is to: Stress the CPU as much as possible)

[OmegaMerc]

how long of a burn have you used, and im sure this is gonna come up, so ill go ahead and throw it in first

A64 cpu's cannot be burned in, whats actually happening is that the thermal compound between the IHS and the core takes at or around 200 hours to get up to operating thickness to be able to make good enough contact with the core to make someone think they've actually burned in their cpu to a higher lever when actually all youve done is break in the thermal compount..

Im in no means downing your proggie or ur overclock in anyway dude, I have actually downloaded it myself, I just knew this would eventually come from someone here because this has been a hot topic here for a long time, I just wanted to beat someone else to the punch :)

Thank you for explaining and settling this claim, was getting bothered of not finding an answer.

[[cTx]Philosophy]

will give it a try though, if it heats better than prime its definately a good way to break in the thermal crap amd uses..

Currently rosetta does its best to try and acheive the same efforts for me :P.. Alas to no avail, still @ 2.7

[NickS]

I'm running this on the School Dell ATM, see how loud I can get it to be ;)

Nick

[Jochenp]

@ Philosophy
Now what if I tell you I ran my cpu for a month without thermal compound (laziness), and I still got a voltage drop and mhz increase burning in? (going from 1.62v to 1.56v and from 2450mhz to 2520mhz).

Jochenp

[Jochenp]

Oh, and the "you've broken the TIM under you IHS"-argument won't work either :p: : I have my chips naked
Jochenp

[r3w4]

how long of a burn have you used, and im sure this is gonna come up, so ill go ahead and throw it in first

A64 cpu's cannot be burned in, whats actually happening is that the thermal compound between the IHS and the core takes at or around 200 hours to get up to operating thickness to be able to make good enough contact with the core to make someone think they've actually burned in their cpu to a higher lever when actually all youve done is break in the thermal compount..

Im in no means downing your proggie or ur overclock in anyway dude, I have actually downloaded it myself, I just knew this would eventually come from someone here because this has been a hot topic here for a long time, I just wanted to beat someone else to the punch :)
Small changes in thermal conductance don't account for the large changes in overclocking stability.

[NickS]

Prime 95 heats my CPU up 2*C more than this?

Nick

[AgentGOD]

For best effect... use 4 threads for each core of your CPU. If it only has 1 core, use 4 threads, if it has 2, use 8.

[r3w4]

I get a "this program failed to initialize properly" error every time I try to run it.

[AgentGOD]

You need Microsoft .NET Framework.

Get it here (http://www.microsoft.com/downloads/details.aspx?FamilyID=262d25e3-f589-4842-8157-034d1e7cf3a3&displaylang=en).

[r3w4]

I tried it with 8 threads. It gives off about the same heat as Prime95. Nothing compares to CPU Burn which gives off 6C more than Prime95.

[Jochenp]

try toast or k7burn, it's 4c higher than cpu burn for me

[Fred_Pohl]

http://www.ultimate-filez.com/xstress.jpg

Download Link:
Here (http://www.ultimate-filez.com/XStress.zip)

Compared to StressCPU this is a joke! OTOH, since StressCPU uses SSE and the Gromacs scientific engine, it can potentially crash any pre-revision E Athlon so this is more AMD compatible.

On a positive note, running 2 threads of X-Stress on my HT P4C lowered my cpu temp by 4C compared to running a single thread of F@H! Maybe you should market this as an alternative to system idle process?

[Jochenp]

Hey, don't bash other people's software
He did a very good job IMO

[fAlCoNNiAn]

used it liked it, but i think alot more people have to post their results in order for people like me to switch to this, as opposed to using prime95 as my main testing. i like the program though, great job!

[Fred_Pohl]

used it liked it, but i think alot more people have to post their results in order for people like me to switch to this, as opposed to using prime95 as my main testing. i like the program though, great job!

All I can offer is that this stress program actually lowered my cpu temp by 3-4C compared to it's normal state running one instance of FAH on my HT cpu. Running FAH and encoding a DVD, as I often do, produces cpu temps 5-7C higher than running this program alone. Running Prime95 produces temps only 2-4C lower than FAH, depending upon which FAH engine is running. Obviously Gromacs(SSE) taxes the cpu more than Tinkers (no-SIMD) and DGro(SSE2) taxes it more than Gro or Tinkers.

[Jochenp]

Are you sure that you ambient temp hasn't changed 4-5 degrees?

[Fred_Pohl]

Are you sure that you ambient temp hasn't changed 4-5 degrees?

Absolutely. Although I doubt that my measurements are any more accurate than +/- 2C.

[pete5990]

Try to make the program as efficient as possible. Then it'll be doing more calculations instead of running useless code, and it will heat the CPU up more.

[Xeon th MG Pony]

it raised my temps 1C over idle with my fan on low, it raised the chip set by 2C though over idle.

BEFOR:
CPU 0= 34
CPU 1= 35

AFTER
CPU 0= 35
CPU 1= 35

Dual Xeon Prestonies 2.8GHz @ 3.2GHz water cooled with the MCW6002, arctic Silver Ceramique (Aplied over 3 years ago), MCP600 on a dual 120mm Heater core, with a 24V fan running at 12V (Half speed)