[Info-vax] Comment on the future of OpenVMS

[Neil Rieck]

On Oct 26, 1:37 pm, JF Mezei <jfmezei.spam... at vaxination.ca> wrote:
> Neil Rieck wrote:
> > Update: I just checked the seminar notes and BL870c was also covered.
> > This single blade supports 4-CPU sockets so I'm assuming that a future
> > version of this card will support up to 16 Tukwila-based cores
>
> Question then becomes: how much later after the 8086, will IA64 get the
> hardware and software to suppport such blade configs.
>
> quickpath + quad core has been available now for over a year on the
> 8086. IA64 with same features won't be available for another few months
> at least.

Good question. IIRC there was never any mention of quickpath. Gaitan
D’Antoni mentioned that he would answer emails so why don't you shoot
him a note (his email address is in the leacture 5 and 6 slides)) and
post his response here.

> And while I understand the usefulness of blades in a windows environment
> where you need multiple separate instances of Windows to run multiple
> applications, I still don't really see the usefullness of blades in an
> enterprise OS where you would want separate systems for redundancy
> purposes and where you don't need multiple systems to run multiple
> applications.

I am probably the wrong person to be answering this question, but
here's by 2-cents worth: Blades provide a more compact way to install/
maintain CPU equipment in either a computer room or the factory floor.
The cards are all hot-swappable (compare this to my AS-DS20e where you
need to open/maintain it in a traditional way; it's really tough
adding memory DIMMs because you are working on a dark mobo at the
bottom of the chassis). They also seem to support virtualized network
connections so that adding more blades to the rack doesn't require
adding new network connections.

Maybe someone a little more knowledgeable should pick up where I just
left off :-)

NSR