760GM-E51 locking up on heavy HDD usage

Question

Please see hardware information in sig.
Everything was running smoothly for quite some time with the WD and Seagate SATA HDDs using IDE emulation (no Hitachi at this point). Over the last month the system has been locking up unexpectedly, no useful error reports in Windows to suggest the cause. The main factor appeared to be transferring large quantities of data between those two drives. So I look arround, found out about the WD issues w.r.t sleeping after only 8 seconds and TLDR, increased the sleep time to 5 minutes, the drive didnt support TLDR so I could not enable it.
For my own personal needs, and to aid the fault-finding I purchased the Hitachi 2tb drive, which I installed last night. Then moved both SATA drives from IDE emulation to AHCI, incase the TLDR difficulties were causing the SATA controller to drop the WD drive unexpectedly if it took too long to respond - which with IDE emulation could cause serious issues. Moved some data between the WD and Seagate (~10gb, which was normally enough to trigger the problem), nothing happened, which was a good sign. I then left the Hitachi drive formatting, and during this time decided to stream a video over the WiFi connection. A few minutes in the video stops, the SAMBA shares are disconnected, and the machine no longer responds to ping. At the machine there is no response to keyboard or mouse, and the HDD activity light is solid red - the lock-ups are still occurring, exactly sa before. I should clarify that at this point if the TFT was not on, switching on the TFT tends to result in a large gray image, at supposedly the correct resolution.
I have run Prime95 for many hours, with seperate instances for all 4 cores (+ correct affinity settings), at both "In-place large FFTs" (max heat & power consumption), and Blend (max ram tested), with no problems. I believe I have also run memtest, but can double-check if necessary. I have also used the machine extensively for numerical simulation with no problems. It would seem that if this was a CPU or power issue there would be other side-effects, and more obvious crashes. I am left stuck with the situation that I believe it is either one of the harddrives behaving badly - and the SATA controller handling this inappropriately, or a chipset issue.
I do have a byte-for-byte copy of the OS, and could install it on the hitachi drive, but frankly something more light-weight might be preferable for this fault-finding exercise. Are my assumptions so far appropriate? How should I proceed? Please advise.
Thanks in advance

Answer 1

Quote from: Bas on 03-July-11, 02:33:33
You really do not know how memory works do you?
Please dont be condescending.
Quote from: Bas on 03-July-11, 02:33:33
The memory that is being used in Desktop is unbuffered and non-ECC, as such it takes 1 unstable cell to cause an BSOD instandly.
Ofcourse it depends a bit on the instruction, but be sure you will see BSOD's every time you hit that cell.
This does not match my own experience, the experience of others I know who do a fair degree of overclocking (and hence have to avoid these instabilities), or match my understanding of how the average OS works. There is no reason why a BSOD should occur when a single address comes back with the wrong value - that is utter nonsense, that is also the reason why you dont see BSODs whenever a buffer overflow occurs. Corrupting areas of memory by screwing up a pointer in C has basically the same effect as simulating a single error from unstable memory, and that is highly unlikely to cause a BSOD. If you start overwriting data in kernel mode there is plenty of reason to believe the OS will crash, but in user mode? it is unlikely to have a major effect at all. It is only because failing RAM tends to error at such a high rate that these kinds of crashes occur, (equivalent to eventually addressing something important and the usual error-handling failing). We are not talking about chips which are consistently returning bad data, we are talking about instability, the rate of that instability has a massive impact on its outcome.
Quote from: Bas on 03-July-11, 02:33:33
It's not a massive over-generalisation, as these are not servers that have ECC (Error-Correction-Code), buffering or registering of memory.
Bad memory in desktops is a hit -> wrong data -> BSOD thing
Server memory is different: hit -> ECC-check -> bad cell -> rewrite due to ECC corrective code -> placed in new cell -> deliver correct data.
Your system will NOT run for hours without a BSOD if your memory is bad.
Memory is NOT your problem, 100% sure of that.
This makes no sense, I have already explained that Memtest failed, by definition this means that an address in memory returned an incorrect value. According to your reasoning I should already be experiencing BSODs constantly. This is clearly not what is happening. Having said that, the memory returning the wrong value is clearly incorrect behaviour, and even if it were not causing the lock-ups, this is evidence of the system being unstable. Im not sure why you are so forcibly trying to argue with this.
Quote from: Bas on 03-July-11, 02:33:33
As for Smart, I can show you plenty drives where Smart reports nothing and the drive is bad, I can also show you drives where Smart reports it's bad and the drive is trouble free.
Smart is a dumb as a duck, but for companies useful as when it tells you it's bad it's saver to replace the drive before it does go bad.
Smart is not a trustworthy mechanism, it consist of counters that try to predict the future, nothing more nothing less.
Test your harddisks with the manufacturer tools, pretty sure it will find the bad one....it is NOT your memory.
I referred to SMART because that is what cronnin was making me check. Im fully aware that SMART is not a definite indicator of the health of the harddrive, and I never claimed it was, that was not in response to your comments.
I am trying my best to be cordial, but am unhappy with the tone and accuracy of some of these posts, which makes it very difficult to counter your argument in a civilized manner. I dont want this thread to devolve into a heated argument and I dont have any personal problem with you. Can we please have another moderator check over these posts and clean up this situation? It seems to me that if we still want to discuss this memory matter, it would be better suited to PM so as not to clutter this thread. Do you think that is a sensible course of action?