Jump to content

Possible cause for failing FPU stress test? (HP Pavilion Elite HPE-498p + Core i7-870)


Arnauld
 Share

Recommended Posts

HI everyone,

I have been through the forum and there are interesting post everywhere.

My PC is crashing when I am encoding videos, so I started investigating.

I noticed my CPU temperature were going above 70's and thought the was causing the PC black out.

I change my CPU fan for a Cooler Master Blizzard T2. Now in stays far below 55C.

My PC is still not passing the AIDA64 stress test. So I isolated my test.

CPU Test: okay

FPU test: immediate crash after I click on 'start' button. lol

Going through the forum I identify it could come from Power supply unit.

I therefore changed my 460W PSU for a 760W Seasonic X series 80 Plus Gold.

Click on 'start' button -> immediate crash (PC restarting).

I tested memory: all okay.

What would you suggest I do next?

I have a HP Pavilion Elite HPE-498p: i7-870, 2,93MHz, MSI MS-7613, bus 133MHz, 8Go DDR3, NVidia GTX 460.

Thanks in advance.

Arnauld

  • Like 1
Link to comment
Share on other sites

The report is now attached.

Please note that I have now slightly overclocked the CPU (3.2GHz instead of 2.93GHz).

I did it with AIDA, and I have not managed to set it back to normal.

Here is the product spec as original:

http://h10025.www1.hp.com/ewfrf/wc/document?docname=c02542167&cc=sg&destPage=document&jumpid=reg_r1002_sgen_c-001_title_r0002&lc=en&tmp_docname=c02542167

Again, the only hardware I changed are the CPU cooler and the PSU.

Regards.

Arnauld

Report.txt

Link to comment
Share on other sites

Thank you. If you have a chance to borrow two identical DDR3-1333 modules from a friend, then it would worth checking out whether it makes a difference. If you cannot do that, then it may worth playing around with the memory modules, because there may be one or two weak ones amongst the 4 modules you've got installed. Try removing two, and check out if it changes anything. If not, then try replacing the two installed modules with the ones you removed previously, etc. Make sure to have the 2 modules always placed in the DIMM slots that makes the chipset operate optimally (dual-channel; first DIMM slots usually).

I apologize for not being able to provide better ideas on troubleshooting. Remotely it's usually quite difficult :( But what you did so far (installing a better CPU cooler and a new PSU) were both good steps in trying to hunt down the cause of the hardware issue.

  • Like 1
Link to comment
Share on other sites

Hello and thanks for your support, you do not need to apologize, you're providing great advise,

I will try out the memories, and before I will try to find several memory test programs.

How did you come to identify that 1 or 2 out of 4 would be weak? Statistical reasons?

I am expatriate in Singapore and do not know have g33ks friends here.. :P

So I guess I'll have to buy new memory bars, and as the PC do not like mixed manufacturers, I'd better be sure it's going to be the right choice.

Thanks again for your support.

As you mentioned, there are not so many sources for failure.

Fan, PSU and memories are still hardware origin.

Is there a possibility that the cause of failure is a software origin?

If yes, how could I search for it?

Cheers.

Arnauld

Link to comment
Share on other sites

How did you come to identify that 1 or 2 out of 4 would be weak? Statistical reasons?

There's a very slim chance that more than 1 memory module is failing, and an even slimmer chance that more than two. So yes, it's pure statisics :)

And quite frankly, I personally am not fond of using more than two memory modules in a mainstream desktop motherboard (ie. one that has 4 DIMM slots and uses a dual-channel memory controller). So if there's any possibility of switching to just 2 identical memory modules, that alone could fix the issues in many cases. I'm not saying in your particular case that it is definitely the solution, but if you could just remove 2 memory modules, it's a test that worth the extra time and efforts. Removing two of them for a test doesn't cost any $$ at least ;)

So I guess I'll have to buy new memory bars, and as the PC do not like mixed manufacturers, I'd better be sure it's going to be the right choice.

Before you continue spending more $$ on your system, make sure to try the "remove half of the memory modules" test run. Also, you may want to try playing around with memory timings as well, _if_ your BIOS Setup allows altering timings. Upping the Command Rate (CR or tCMD) from 1T to 2T might help as well to stabilize the memory controller.

Is there a possibility that the cause of failure is a software origin?

If yes, how could I search for it?

I'm afraid software issue, operating system issue and configuration issue are always a possibility. A wrong device driver, a wrong service could cause all sorts of issues, not to mention potential malware that has already infected the computer. And then there's the slim chance that an installed software or driver getting "freaked out" when AIDA64 puts extreme stress on the system. Please note that the "FPU" subtest in AIDA64 System Stability Test puts such a heavy workload on the system that is quite unusual, and many systems fail to run that test for more than a few minutes. The "FPU" subtest is definitely the worst case scenario a computer could go through, but a properly built and properly configured PC has to pass that test to be considered a stable system ;)

So if you could save all your data, documents and settings, and have the possibility to completely reinstall Windows and all software from scratch, then it would definitely worth a try. Having a clean Windows installation could improve the response time of the system as well.

  • Like 1
Link to comment
Share on other sites

  • 1 month later...

Hi again,

I am back. :-)

I did further test with AIDA:

1. All CPU protocol passed with success: Queen, PhotoWorxx, ZLib, AES and Hash

2. With regards to FPU tests:

  • - protocol Julia: passed
  • - protocol Mandel: passed
  • - protocol SinJulia: passed
  • - protocol VP8: badly failed

Can you help me understand what is specific to the VP8 FPU protocol that would make my i7 fail?

And most importantly, is there anything I should do to pass it?

Thanks.

Arnauld

Link to comment
Share on other sites

Click on 'start' button -> immediate crash (PC restarting).

This is what I mean.

To be more precise, the PC shuts down immediatly, just like a PSU failure, and then restart.

That is why, one of the first things I did, was to upgrade my PSU (460w to 760w) but with no success.

As a result, I get no BSoD or score. It just shuts down and restart.

Link to comment
Share on other sites

If any of the benchmarks or the System Stability Test causes the system to power off, then it's definitely a hardware flaw that needs to be addressed. If your system is overclocked, then try to restore all settings to their factory default values. Also, updating the motherboard BIOS to the latest one may also help. If those wouldn't help, and you already swapped the PSU, then it could either be a memory (RAM) issue or a motherboard failure. It's quite difficult to diagnose such issues remotely ;)

Link to comment
Share on other sites

  • 3 weeks later...

Just to keep you all updated and also in case this could help someone, I will explain below how I have managed to solve this issue.

After contacting HP support specialist and going through all the 'fixes'I had done before, he could narrow the possibilities. It actually come from the hyper-threading. My PC is with an Intel i7, and hyper-threading was activated. Since I dis-activated it from the boot settings, the PC works fine, I managed to go through all FPU test and run all my games including Far Cry 3.

Hope this help others.

Cheers.

Arnauld

Link to comment
Share on other sites

  • 2 years later...
Hi everyone, i have same problem almost 8 months (sinjulia tests freeze pc). It wasn't memory issue. I found solution for me: i just lowered voltage for CPU from 1.20 to 1.15 (also CPU term. from 1.20 to 1.15) and changed RAM voltage from AUTO to Normal. After all - no more freezes and stucks (passed all FPU tests).

 

Hope it will help somebody.

 

My specs:

 

OS: Windows 8.1 Home x64

CPU: QuadCore Intel Core 2 Quad Q8200

Motherboard: Gigabyte GA-G41M-ES2L

Memory: DDR2 4gb 800 (Transcend)

GPU: Gigabyte GeForce GTS 450

HDD: Seagate 500gb Sata II (5400rpm)

Link to comment
Share on other sites

Hi everyone, i have same problem almost 8 months (sinjulia tests freeze pc). It wasn't memory issue. I found solution for me: i just lowered voltage for CPU from 1.20 to 1.15 (also CPU term. from 1.20 to 1.15) and changed RAM voltage from AUTO to Normal. After all - no more freezes and stucks (passed all FPU tests).

 

Hope it will help somebody.

 

My specs:

 

OS: Windows 8.1 Home x64

CPU: QuadCore Intel Core 2 Quad Q8200

Motherboard: Gigabyte GA-G41M-ES2L

Memory: DDR2 4gb 800 (Transcend)

GPU: Gigabyte GeForce GTS 450

HDD: Seagate 500gb Sata II (5400rpm)

Thank you for posting your solution.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

×
×
  • Create New...