Windows Vista X64 bsod
in [64 Bits Windows]
|
Prev: Dual-booting Windows 7 and Ubuntu 9.10 Karmic Koala
Next: Problem with the Private Character Editor and Vista 64 Bit
From: R. C. White on 16 Nov 2009 22:59 Hi, Richard. As I've reported before - and you may have seen - sometimes RAM that passes all the tests can still fail. When I built my computer at the end of 2006, I put in 2 x 1 GB OCZ PC-6400 DIMMs. That ran Vista Ultimate x64 just fine for a year. Then I added 2 more identical sticks and it ran with 4 GB for about 6 months. Then, last year, the computer started producing BSODs. Since the Stop codes varied randomly, I knew the problem was in hardware, not software, and I suspected memory problems. The RAM passed all my tests and those run by my local computer shop, but we got flaky results with one of the newer sticks was installed. I ran with 3 GB for a few months, then I contacted OCZ and the tech had me return BOTH new sticks so that they could test them as a matched pair. OCZ RMAed both sticks - and I've had no problem in the several months since then. So, even when RAM passes all the field tests, there might still be a problem that only the manufacturer can fix. RC -- R. C. White, CPA San Marcos, TX rc(a)grandecom.net Microsoft Windows MVP Windows Live Mail 2009 (14.0.8089.0726) in Win7 Ultimate x64 "Richard Urban" <richardurbanREMOVETHIS(a)hotmail.com> wrote in message news:e1fx9AzZKHA.196(a)TK2MSFTNGP05.phx.gbl... > The hardware tester stress tests the RAM at a higher operating voltage and > a lower operating voltage - something that can not be done while the RAM > is in the computer. > > -- > > Richard Urban > Microsoft MVP > Windows Desktop Experience & Security > > > "John Whitworth" <sexyjw(a)gEEEEEEEEEEEmail.com> wrote in message > news:OpswRcrZKHA.5608(a)TK2MSFTNGP05.phx.gbl... >> >> "Richard Urban" <richardurbanREMOVETHIS(a)hotmail.com> wrote in message >> news:OJR0Q0HZKHA.1640(a)TK2MSFTNGP06.phx.gbl... >>> Memtest86 is next to useless. >> >> Memtest86 has been able to confirm my suspicions about dodgy RAM on each >> occasion that I've been unlucky enough to acquire it. That's good enough >> for me. It may not be super accurate, but it is a good start, and it will >> probably help 80% of those looking to isolate issues. Your post is just >> trying to get people out of the self-fix habit. Sure, if they don't find >> a fault after Memtest86, take it to a 'pro'. But if it enables them to >> prove there is a problem without incurring costs, why should you deter >> them? >> >> Not only that, but could I guarantee that my PC is running with the same >> speed and voltage settings as the hardware tester? I can with Memtest86. >> >> JW
From: John Whitworth on 17 Nov 2009 03:30 "R. C. White" <rc(a)grandecom.net> wrote in message news:#XnoNpzZKHA.1652(a)TK2MSFTNGP05.phx.gbl... > Hi, Richard. > > As I've reported before - and you may have seen - sometimes RAM that > passes all the tests can still fail. > But even so, it's a simple and cost effective step to start with Memtest86. JW
From: John Whitworth on 17 Nov 2009 03:33 "Richard Urban" <richardurbanREMOVETHIS(a)hotmail.com> wrote in message news:e1fx9AzZKHA.196(a)TK2MSFTNGP05.phx.gbl... > The hardware tester stress tests the RAM at a higher operating voltage and > a lower operating voltage - something that can not be done while the RAM > is in the computer. Surely by testing it at different voltages, it is going to have different results to when it is in the computer. So Memtest86 doesn't stand a chance of picking that up - and why would it? Memtest86 is simply trying to see if there is a problem - in that computer - with the current settings. The hardware stress tester sounds great to me if you want to make sure that memory is going to survive extreme temperatures or overclocking etc. But whichever way you look at it, Memtest86 is an ideal first check, screening out a lot of simple-to-find problems. JW
From: Richard Urban on 17 Nov 2009 22:30 How about "normal" voltage variations from the computers power supply. I have seen few darn computers with an honest 12 volts, 5 volts and 3.3 volts. Many supplies are +- 10% of the stated voltage. Marginal RAM may begin faulting under those conditions. If you are lucky the voltage will be low/high when you run the software memory test programs and you will pick up on the condition. -- Richard Urban Microsoft MVP Windows Desktop Experience & Security "John Whitworth" <sexyjw(a)gEEEEEEEEEEEmail.com> wrote in message news:OGpgbC2ZKHA.5472(a)TK2MSFTNGP02.phx.gbl... > > > "Richard Urban" <richardurbanREMOVETHIS(a)hotmail.com> wrote in message > news:e1fx9AzZKHA.196(a)TK2MSFTNGP05.phx.gbl... >> The hardware tester stress tests the RAM at a higher operating voltage >> and a lower operating voltage - something that can not be done while the >> RAM is in the computer. > > Surely by testing it at different voltages, it is going to have different > results to when it is in the computer. So Memtest86 doesn't stand a chance > of picking that up - and why would it? Memtest86 is simply trying to see > if there is a problem - in that computer - with the current settings. > > The hardware stress tester sounds great to me if you want to make sure > that memory is going to survive extreme temperatures or overclocking etc. > > But whichever way you look at it, Memtest86 is an ideal first check, > screening out a lot of simple-to-find problems. > > JW
From: David B. on 18 Nov 2009 12:23
Or you can just plug the supply into a power supply tester which is far cheaper than memory testing hardware. Why would you be worrying about the RAM passing or failing if your power supply is so far out of spec? -- -- "Richard Urban" <richardurbanREMOVETHIS(a)hotmail.com> wrote in message news:%23P%23%23N%23$ZKHA.196(a)TK2MSFTNGP05.phx.gbl... > How about "normal" voltage variations from the computers power supply. > > I have seen few darn computers with an honest 12 volts, 5 volts and 3.3 > volts. Many supplies are +- 10% of the stated voltage. Marginal RAM may > begin faulting under those conditions. If you are lucky the voltage will > be low/high when you run the software memory test programs and you will > pick up on the condition. > > -- > > Richard Urban > Microsoft MVP > Windows Desktop Experience & Security > > > "John Whitworth" <sexyjw(a)gEEEEEEEEEEEmail.com> wrote in message > news:OGpgbC2ZKHA.5472(a)TK2MSFTNGP02.phx.gbl... >> >> >> "Richard Urban" <richardurbanREMOVETHIS(a)hotmail.com> wrote in message >> news:e1fx9AzZKHA.196(a)TK2MSFTNGP05.phx.gbl... >>> The hardware tester stress tests the RAM at a higher operating voltage >>> and a lower operating voltage - something that can not be done while the >>> RAM is in the computer. >> >> Surely by testing it at different voltages, it is going to have different >> results to when it is in the computer. So Memtest86 doesn't stand a >> chance of picking that up - and why would it? Memtest86 is simply trying >> to see if there is a problem - in that computer - with the current >> settings. >> >> The hardware stress tester sounds great to me if you want to make sure >> that memory is going to survive extreme temperatures or overclocking etc. >> >> But whichever way you look at it, Memtest86 is an ideal first check, >> screening out a lot of simple-to-find problems. >> >> JW > |