Low Confidence Level - what now?

[epags]

I am new to using Memtest 86. So I made a couple of mistakes in interpreting the data so now I am turning to the forum expertise for help.

My system is: a computer shop’s Custom Build
Processor - 1 x Intel Core i5 750 / 2.66 GHz - LGA1156 Socket - L3 8 MB

Intel Desktop Board DP55WB Media Series - Motherboard - micro ATX - iP55 - LGA1156
Socket - Serial ATA-300 (RAID) - Gigabit Ethernet - FireWire - HD Audio (8-channel)

Kingston Value RAM - Memory - 8 GB – 2-( 2 x 2 GB ) - DIMM 240-pin - DDR3 - 1333 MHz /
PC3-10600 - CL9 - 1.5 V - unbuffered - non-ECC

I have been having occasional BSOD failures for over a year. The original company who built the computer has fiddled with it and could not fix the problem. They thought it might be the graphics board and I changed it. BSOD problem continued.

I found Memtest86 and ran the test. I jumped the gun again and stopped test after one hour when Error Confidence Level was 81% and tests 3, 4,and 5 each showed 65,535+ errors. Photo https://www.dropbox.com/s/kfp1udbujcugyhm/100_5324.jpg
Of course I hadn’t check this forum so bought new RAM and replaced all modules.

Today I replaced all the Kingston memory with new Crucial Memory 8 GB – 2-( 2 x 2 GB ) - DIMM 240-pin - DDR3 - 1333 MHz PC3-10600 - CL9 - 1.5 V - unbuffered - non-ECC (A slightly costly thing to do based on my rush to solve the problem)

Ran test again, this time letting it run through a full iteration. Note that the confidence level has dropped to 36% and here is a screen shot: https://www.dropbox.com/s/ba9iyw2uxphlcko/100_5329.jpg

Where do I go next with this problem?

[David (PassMark)]

Are the BSOD errors you get random or it is always the same crash. Random BSOD more likely point to RAM problem. A consistent BSOD error of the same type is more likely to be a software fault.

MemTest is reporting a HUGE number of errors across a wide range of memory addresses.

The Confidence Level isn't a percentage. Don't ask me exactly what it is as I don't actually know. Clearly we need to document this better. Cbrady if he is around might have some background information on this.

But the number of errors is the real problem.
If you had a spare CPU or motherboard I would suggest swapping them out.

Also something interesting. You claim in your post above you have a Core i5 750. But in the screen shot you have a Core i7 860 ?

[epags]

Low Confidence continued

Thanks for the quick come back.

You said: Are the BSOD errors you get random or it is always the same crash. Random BSOD more likely point to RAM problem. A consistent BSOD error of the same type is more likely to be a software fault." Here are the last three screen shots of BSOD:
https://www.dropbox.com/s/6owlmpoct5pt2z7/100_5315.jpg
https://www.dropbox.com/s/cbrnofvxh8fana0/100_5316.jpg
https://www.dropbox.com/s/1ejxh2sserw1tju/100_5322.jpg


You also said: "Also something interesting. You claim in your post above you have a Core i5 750. But in the screen shot you have a Core i7 860 ? "

You are right....I was quoting the specs the builder had given me. I did run a scan and the real build is a Core i7 860.

Wonder what else was not to the specs???

You said: "The Confidence Level isn't a percentage. Don't ask me exactly what it is as I don't actually know. Clearly we need to document this better. Cbrady if he is around might have some background information on this."

I thought I had to have a score of 100% or better for the error to be the RAM. Hope Cbrady can help in this area.

Again...thanks

[David (PassMark)]

The BSOD errors seem random enough. Decreasing the chance of it being a device driver bug.

As you have already changed the RAM, the next most likely hardware cause is a bad MB. You could try testing just 1 memory slot at a time. But even if you discovered it was just 1 slot causing the problem, you might still want to replace the entire MB.

[epags]

Thanks: Looks like any fun I scheduled for today, Sunday, is going to be cancelled while I go slot to slot testing.

Now I am reading up on how to change the mother board....ugh.

Thanks again.