I'll always remember the phone call from the chip engineer who knew not
much of anything about software or systems. He did know that memory at
that point was largely virtual and knew that data was paged in and out
to disk. He had called me to find out how long data was typically
resident on-chip before being paged in or out. He had this notion of
the time always being measured in seconds or minutes which would have
meant a low probablility of a cosmic ray strike. When I told him that
some code and data were fairly permanently resident and it could be
hours, days or weeks or more he said: "Uh, oh!"
Chuck Norcutt
Tim Hughes wrote:
> I seem to remember reading the normal memory chips suffer from ionizing
> radiation hits, (because
> the number of electrons stored has got to be so small), but they now
> encapsulate/cover the chip
> surface to minimise the effect.
>
> tim Hughes
>
>
>>that too. But I do not know if those bits will eventually fade off into
>>the void sitting in the closet or if they're subject to flipping from
>>cosmic rays.
>>
>>I do know that cosmic rays were a worry to at least one of IBM's memory
>>engineers when chip density hit 64K bits about 30 years ago. But that
>>problem was obviouly solved or else maybe didn't exist in the first place.
>>
>>Chuck Norcutt
>
>
>
> ==============================================
> List usage info: http://www.zuikoholic.com
> List nannies: olympusadmin@xxxxxxxxxx
> ==============================================
>
>
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|