Re: Chipsandcheese article on the CDC6600

On Tue, 23 Jul 2024 00:20:46 -0000 (UTC)
Lawrence D'Oliveiro <ldo@nz.invalid> wrote:

On Mon, 22 Jul 2024 13:08:27 +0300, Michael S wrote:
 

At the end, the influence of 6600 on computers we use today is
close to zero. 

 
He pioneered pipelining

It (6600), not he (Seymour).
No, it didn't.
If instead of 'it' we'd talk about 'he' then the first Cray-Thornton
pipelined computer is 7600. But by then pipelining was hardly new.
7600 can be arguably credited for  "pipelining done right", but not as
a pioneer in that area.

and multiple function units.

Multiple functional units existed before. The special thing about 6600
was that it had ALOT of them. Having a lot of non-pipelined functional
units after 1-wide or even 2 or 3-wide front end sounds like
architectural dead end.

He went on to
pioneer vector processing (long vectors, not the short-vector SIMD
stuff that infests CPU designs today).

I am talking about 6600, the computer, not Seymour Cray, the person, or
Cray-Thornton, the team.
Cray, the person was innovative and influential.
6600, the computer was innovative and not influential in the long run.

He was always very
conservative in the fabrication technologies he adopted, but he was
brilliant at pushing them to their limits.