3 Mar
2008
3 Mar
'08
6:51 a.m.
Maybe a stupid solution, but couldn't the over heating be solved by
lowering the clock... if possible?
My theory after following this thread is that the overheating was caused
by a once borderline processor running too hot for too long causing
migration of dopants in the transistors and increasing the static power
drain.
I once had a x86 server that I overclocked for a year (calculating prime
numbers) running at full load until it started to become erratic. It was
rock solid for a year and then I lowered the frequency to the marked
frequency and five years later it was still running rock solid.
I finally had to upgrade when I was /. by Ericsson after my web address
was publicised on the intranet of Ericsson on a friday afternoon. Load
was at 35 and rising fast. :-D
Good luck!
/G?ran
der Mouse wrote:
> Is the
[uV2k] FPU socketed?
>
It's soldered down and not removeable with
out the right tools for
handling high lead count surface mount.
Me I'd pull a old MVII cpu card and do the
remove and replace of the
FPU and CPU. MicroVAXIIs with good cpus and FPU and bad Qbus
interfaces are not uncommon.
The uVII and the uV2k use identical FPU and CPU chips, even down to the
packaging? That's definitely a tidbit worth remembering.
I don't think I have a suitable uV2 donor card just now, though. :(
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B