Newsgroups: comp.lang.lisp.mcl,alt.sys.pdp10
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!ix.netcom.com!netcom.com!alderson
From: alderson@netcom.com (Richard M. Alderson III)
Subject: Re: XEROX 1100, PDP-10 performance...
In-Reply-To: e@flavors.com's message of 20 Mar 1995 11:17:22 -0500
Message-ID: <aldersonD5vBEG.JE@netcom.com>
Followup-To: alt.sys.pdp10
Reply-To: alderson@netcom.com
Fcc: /u52/alderson/postings
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <v01510100ab9356ed1bf3@[204.5.215.14]>
Date: Thu, 23 Mar 1995 00:44:40 GMT
Lines: 62
Sender: alderson@netcom20.netcom.com

In article <v01510100ab9356ed1bf3@[204.5.215.14]> e@flavors.com
(Doug Currie, Flavors Technology, Inc.) writes:

>At 1:15 AM 95.03.16, JasonStowe wrote:

>>reference to a "SUMEX, PDP-10, with a KI-10 uniprocessor 256k core memory"

>These were microcoded machines. The KI-10 and KL-10 were designed by Bob Reid,
>and used in the popular DecSystem 2020. I was lucky to work with Bob in 1977,
>right after he did these CPUs, on a new CPU for Gould; he did the hardware, I
>did the microcode. Based on what I remember about the technology of those
>machines, the microcode cycle was 100 nsec (ECL) to 250 nsec (TTL). A user
>level instruction might take four to twenty (or more for floating point or
>multiply/divide) microcycles. So the top end for these "fire breathers" was
>about 1 MIPS even with aggressive pipelining and caching. The VAX was also
>being developed at this time. The fastest first generation Vaxen were also
>about 1 MIPS.

The KI10 was not microcoded; the KL10 was.  Neither was used in the 2020, which
was a KS10 processor (a microcoded implementation based on 2901 bit-slice
technology).

The KI10 precedes the DEC-20 line by 5 years; the first DEC-20 was the KL10.

The DEC-10 was originally the predecessor to the KI10, the KA10; later, the
KI10, KL10, and KS10 were all used in DEC-10 products.

The KL was roughly a 1MIPS processor; the KI was about half as fast, and the KA
about two-thirds of the KI.

None of these machines was pipelined; later versions of the KL had a cache, but
not the first.  The KS may have had a cache (I doubt it); neither the KA nor
the KI did.

The first DEC 36-bit system, the PDP-6, came out in 1964.  The KA arrived in
1968, the KI in 1970.  The KL came out early in 1975; the KS was a 1979 entry.
The first VAX didn't come out until 1977.

Other manufacturers (Foonly, Systems Concepts, XKL Systems) have manufactured
PDP-10 clones.  No one has ever had the bad taste to bother to duplicate the
VAX.

>Gabriel has a book on "Performance and Evaluation of LISP Systems" (MIT
>Press?) which will have data on these and many other systems. I don't have my
>copy handy, so I can't give you any hard data (and please forgive me if I've
>got the title wrong). If you can't find the book, write me directly for a
>better reference.

Although RPG ran both Portable Standard Lisp on a DEC-20 and MACLISP on SAIL (a
KL10 running a Stanford-local OS called WAITS), neither system is described in
terms of "MIPS numbers."

By the time I got to Stanford, SUMEX had been a KL for years, so the original
poster's data was *long* *LONG* out of date.

None of this is relevant to comp.lang.lisp.mcl.  I apologize for taking up
everyone's time there.  Follow-ups set accordingly.
-- 
Rich Alderson		[Tolkien quote temporarily removed in favour of
alderson@netcom.com	 proselytizing comment below --rma]

Please support the creation of the humanities hierarchy of newsgroups!
