AMD Phenom II X4 955 BE, Gigabyte GA-870A-UD3 with F2 BIOS, GeForce 7600GS, WD 74GB 10000 RPM raptor, WD 320GB 7200 RPM, CoolMax CP-500T 500 watt power supply
If you need anything else, let me know.
What you have there should not be that much of a stress for the PSU. I would run http://www.memtest.org/ though.
I don’t see any CD / DVD / Blu-ray drives listed. Is that all the hardware in your system?
If so, 500W PS should be ok (I guess), as you’re only powering the vid card, MOBO and two disks as far as I can see.
Take a look here to see how much power some devices use: http://www.codinghorror.com/blog/2006/08/video-card-power-consumption.html
For your processor: http://techreport.com/articles.x/16796/12 (about 200 watts or so on average?)
Obviously there is either a memory problem or a lack of power, and it’s looking like the memory.
I’d use either one of these:
Windows memory diagnostics: http://technet.microsoft.com/en-us/magazine/2008.09.utilityspotlight.aspx
memtest86 (link given earlier).
I’m also using a sata Lite-On CD/DVD burner and a Creative X-Fi XtremeGamer sound card.
I updated my BIOS yesterday and ran Memtest86+ v4.10 over night last night for about 10 hours and all 12 passes were successful with no errors. Even before the BIOS update, I had no errors.
It appears my memory is fine according to Memtest86+, which is highly recommended for memory diagnostics. I have ran Prime95 v25.11 for a full hour with no errors. I have also ran OCCT (overclock checking tool) on the linpack benchmark for the full hour with no errors or crashes. Windows 7 64-bit runs just fine in some of the games my video card can run along with browsing in Firefox 3.6.6 with minimal extensions. None of my hardware appears to be faulty or causing any problems as far as I can tell. I don’t know what else I can do to try and find the faulty hardware, if there is any.
Ok this might be a longshot, but I do remember some post by Vlk on the blog that mentioned speed improvements due to some “tricks” the avast tam learned from an Intel team. Now I was just wondering whether maybe, avast tried something “funky” on an AMD cpu and maybe it was not supposed to.
True, given this line of reasoning maybe more people using AMD cpus experience weird crashes, do they ? At least this guy http://forum.avast.com/index.php?topic=61584.0 seems to.
It is very weird that Avast is able to crash Win 7 64bit. Weren’t all drivers in userland now ?
@silviucc:
...that mentioned speed improvements due to some "tricks" the avast tam learned from an Intel team.
IIRC, these speed optimizations were just for hash counting with using new CPU instructions (like SSE instructions set, pipelines), etc. Other changes were in engine, but it had nothing to do with this crash. Wolfman86 sent me some dumps, some of them really points on invalid memory access (e.g. under high CPU load) - we could saw CPU really tried to read the invalid memory (this memory address was also stored in CR2, the internal CPU registry) although program wanted to read a memory block from the different memory location. This must be a HW failure, because you can’t achieve this state programmatically (i.e. can’t be a software bug).
No, certainly not.
Wolfman, can you give us an idea about the temp. of the CPU idle / working?
Idle temps are 39-41c and full load is usually 65-66c, but that’s on the stock fansink. My video card seems a bit hot sitting idle at 49-50c. Not sure what it gets up to on full load, but I can imagine it gets pretty hot.
I noticed in my BIOS, after updating it the other day, it shows 4 cores in the overclock/unlock cores section. Before, it only showed 2 cores. Would that make any difference?
Here are pics of my memory BIOS settings if it may help…
http://img85.imageshack.us/img85/9510/bios1.th.png
http://img39.imageshack.us/img39/2142/bios2.th.png
Note: OCZ said it was ok to use 1.66 volts because 1.65 isn’t available in my BIOS.
What does it do when you select “Auto” for “DDR3 Timing Items” ?
Maybe some more relaxed timings could improve your system’s stability ?
Also, looking at this http://memoryselector.cnetchannel.com/ocztech/result.asp?fk_class=6569&manufacturer_id=781&model_id=941714 I do not see any of the “Black Edition” codes listed
Setting DDR3 timings to auto sets it to everything in the far right column.
This is the official memory support list for my motherboard, and yeah I don’t see any of the OCZ Black Edition kits listed: http://download.gigabyte.us/FileList/Memory/mb_memory_ga-870a-ud3.pdf
How much of an impact do those settings have on the performance of the system (benchmarks, Windows Experience Index…) ?
Maybe it’s worth the tradeoff if those setting help you run stable or you could look with CPU-Z at all the timings the SPD reports as being supported, then manually adjust to the most “relaxed” ones if 11-11-11-30 is causing too much of a performance drop
I decided to test both sticks individually again and one of them gave me errors this time. I tested the other stick in the same slot and it passed 4 times with no errors. I’m starting to think maybe I have bad memory slots…? I don’t know, but I’m going to test each stick individually, again, over night and see what happens.
http://img22.imageshack.us/img22/9995/photo0008nn.th.jpg
http://img23.imageshack.us/img23/4672/photo0010es.th.jpg
Any updates regarding this issue ? What was causing the BSODs, faulty RAM, faulty RAM slot ?
Software does not create hardware issues that simple.
Seems just the memory is faulty and you need to change it.
For the past 48 hours I have tested both sticks, individually with Memtest86+, in memory slot 1 (closest to the cpu) and both have been successful with no errors. So, I either have a BIOS setting that is off or faulty memory slots. I will call Gigabyte tomorrow and find out from them if I really need to RMA or if it’s just a simple BIOS setting that needs to be set/changed.
Cache (memory cache) or overclocking could produce that errors.
Sorry, can’t help further, it’s beyond my knowledge.