[Info-vax] (topic change to rx2600)

fhsjvl at gmail.com fhsjvl at gmail.com
Wed Jun 8 05:34:42 EDT 2016


On Wednesday, June 8, 2016 at 12:38:54 AM UTC+2, MG wrote:
> Op 7-jun-2016 om 08:58 schreef Hans Vlems:
> > Did you encounter problems that the motherboard did
> > not recognize the id's of the new cpu's and then fail
> > to boot?
> 
> I have upgraded a number of rx2600s and rx2620s over the
> years, back when I still owned and ran IA-64 hardware, and
> I never had any problems with it.  (The processors in question
> were "Madison", "Madison-9M" and "Montecito" [so-called 9000-
> series] types, to be specific.)
> 
> The bulk of them even came from non-HP IA-64 systems, no
> lock-ins, cross-platform incompatibilities, etc., fortunately,
> interestingly, I guess, and above all... conveniently.
> 
> The only time I ran into such a problem was with an attempted
> rx2600 System Board swap (I ran into and experienced roughly
> what you described).  I found one cheaply once and I was
> curious if perhaps my System Board was faulty, due to strange
> and erratic behavior.  Later I, like you, swapped the
> 'original' back and used the remainder system board to part
> together a whole new rx2600 from 'scrap'.  Apart from UUID
> mismatches, I was still able to 'handcrank' that rx2600 up
> and running.  I do not remember the details, but it was
> possible, interestingly.  (It wasn't an ideal solution,
> needless to say.)
> 
>   - MG

It should be noted that the zx6000 and the rx2600 are essentially the same system. The zx2000 is different kind of system.

This does not mean everything is incompatible. CPUs can sometimes be swapped indeed. It is also possible to use the Management Processor of an rx1620 in the zx6000/rx2600. And so on.

I assume the "VGA BIOS" entry in the log is irrelevant to the problem of rcyoung. It only says that rcyoung has an NVIDIA Quadro2 graphics card and Steven Schweda has a AMD ATI Radeon graphics card. A zx6000 system with a more recent firmware does no longer give the "VGA BIOS" entry.

It is not entirely correct that the zx6000 "mixed AGP+PCI-X cage backplane riser board" is "essentially of zero use to VMS". When a PCI Radeon graphics card is present in slot 2 or 3 then an AGP Radeon graphics card in slot 1 (the AGP slot) will work correctly with VMS. On the other hand, when a Management Processor (which has the es1000, a basic kind of Radeon) is present, the AGP Radeon card will be disabled automacitally.



More information about the Info-vax mailing list