[Info-vax] Chinese Alpha?

[David Froble]

Michael Kraemer wrote:
> JF Mezei schrieb:
> 
>> Had EV7 work not be stretched and delayed over many years, it would have
>> come out before power consumtion was a big issue, and my guess is that
>> any shrinks and EV8 would have then dealt with it.
> 
> I don't get it what magical design features the Alpha has, which
> would make it still attractive, even if it had been developed further.
> It's a design more than twenty years old, patents would have expired,
> and even by DEC's exaggerated claims it would reach its predicted
> lifetime in five years from now.
> 
>> Had Digital and Alpha survived, I am pretty sure that they would have
>> lead the market with power efficient Alpha, especially once they would
>> have gone multi core.
> 
> I don't know what feeds your pipe dreams,
> but the Alpha never had a good track record in power efficiency.
> In fact, its inability to adapt to the needs of embedded devices
> (with the exception of AXPvme, which failed miserably)
> was one reason for its demise.
> 

When it was being competitively developed, nothing was faster than Alpha.

As has been mentioned elsewhere, some people will pay for performance.

To get to your question, you need to ask, "why was it fast?"  Perhaps because the design 
was better than the competition?  If so, and if there has not been any new designs that 
might be better, then a competitively developed Alpha might still be the fastest single 
processor available.  When talking multiple cores, the "glue" on the chip, memory 
controllers, interprocessor communications, and such in EV7 and EV7z, even with the larger 
die size was doing things that out performed competing CPUs that had continued development 
and smaller die size.

So, if the Alpha could be successfully shrunk down to say 35 nm, or even 22 nm which Intel 
has just released, and with a large on chip cache as the IA-64 has, perhaps it might still 
out-perform anything else available.

Now, that is a mighty big "if", and what it would cost I have no idea, other than it 
wouldn't be cheap.  So, commercially, might it be viable?  I really have no idea.  It 
would depend upon how many might pay for such performance.

My perception is that the majority of people using VMS are not doing so for ultimate 
performance.  If this is so, then the majority just might be satisfied with VMS running 
(and being adequately supported) on x86-64.  Since I feel very strongly that some of the 
reason for VMS losing market share is the perception to some that the HW that supports VMS 
might easily die off, having it on x86 would quash that issue.

Actually, VMS is supported, in emulation, on x86 right now, and the performance appears to 
be satisfactory, else the emulator vendors wouldn't be in business.