Re How many vector registers are useful?

ssimmons@convex.com (Steve Simmons)
Tue, 26 Jan 1993 13:22:15 GMT

          From comp.compilers

Related articles
How many vector registers are useful? kirchner@uklira.informatik.uni-kl.de (1993-01-25)
Re How many vector registers are useful? ssimmons@convex.com (1993-01-26)
Re: Re How many vector registers are useful? billms@corp.corp.mot.com (1993-01-28)
Re: How many vector registers are useful? sanjay@equalizer.cray.com (1993-01-29)
Re: Re How many vector registers are useful? idacrd!desj@uunet.UU.NET (1993-01-31)
Re: Re How many vector registers are useful? billms@corp.mot.com (1993-02-02)
| List of all articles for this month |

Newsgroups: comp.compilers
From: ssimmons@convex.com (Steve Simmons)
Organization: Engineering, CONVEX Computer Corp., Richardson, Tx., USA
Date: Tue, 26 Jan 1993 13:22:15 GMT
References: 93-01-174
Keywords: architecture, vector

On Convex, there are 8 vector registers of 128 elements which are 64 bits
in precision. Obviously, the question is not how many are useful but what
is the optimal number. With 8 vector registers, we still do experience
vector register spills although that is dependent upon the code. Needless
to say, a vector register spill is much worse than a scalar register
spill.


No, I do not know of any published studies that state the optimal number
of vector registers. However, there are costs as more vector operators
are added. These costs are expansion of instruction format, more complex
micro code within the vector unit, more complex chaining analysis in both
the compilers and vector unit, etc... Ultimately, this cost is paid for by
the custumer.


Thank you.


Steve Simmons
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.