Messages in this thread | | | Date | Tue, 23 Apr 1996 08:50:49 +0100 | From | Matthew J Brown <> | Subject | Re: memtest86, built into kernel |
| |
Ulrich Windl writes: > On 22 Apr 96 at 17:29, Matthew J Brown wrote: > > > Basically they say there that with modern SIMMs the MTBF for a soft > > error (ie. alpha particle corruption) is between 12-30 years of > > continual system use. > > Interestingly, this would require to test the RAM for at least 30 > years to be sure. No company will guarantee the MTBF.
No, but that doesn't stop them quoting estimated ones ;) Intel's document extrapolates quoted soft error rates by RAM manufacturers into MTBF, anyway.
As Alan Cox said in a previous message, Intel may be right that soft errors (alpha particle corruption etc) are that rare these days, but their calculations are ignoring all other sources of error, including manufacturing defects.
Intel also have a vested interest in convincing us that parity isn't needed, since they decided not to use it in their popular Triton chipsets.
I don't think that parity gives you that much protection, though, so I'm not convinced that it's worth seeking out systems that support it. ECC may be a different story.
-Matt
| |