25 Apr
1998
25 Apr
'98
2:44 a.m.
How about the CAI system that CDC had (I can't
remember the
name off hand
at the moment - I think my brain is broken - I
couldn't even
remember the
brand name of some wine I wanted to buy yesterday...
Will
chewing on a
floppy disk or two improve things :-). From my limited
exposure
it's likely
that the CDC thingy :-) had a browser style interface.
I saw a CDC demo of that once circa 1974, as I recall it was
called "Plato", used a nice orange plasma display, limited
animation, astronomical price. CDC had grandiose plans it would
be used in schools, like every high school in the country could
afford a $2 million CDC 6600 to support it. Kinda sad actually,
once Cray left CDC they didn't have a clue as to what to do
afterwards, the big iron Cyber 6000s went nowhere, the company
wound up running payroll services (the remnants of CDC are now
known as Ceridian, a batch payroll services company).
Jack Peacock
25 Apr
25 Apr
10:01 a.m.
I think Plato ran on smaller platforms as well.
I do not know if I would say "went nowhere" - the line survived, at least
as an evolution, until the late 1970s. They made CDC money for a little
while, anyway.
Most CDC machines have always been very fast (often the fastest of the
day) and tough, but always a bit behind the times. They refused to upgrade
the line in key areas, most notably their addressing range. In the late
1970s, management finally realized this and the Cyber 180 was born - a
fast mainframe that actually was innovative and modern, but way too late.
I think Seymour Cray was a bit to blame as well, as his machines were also
very fast, but lacking in some areas. The Cray-1s have quite a limitted
memory, and no virtual memory behind it, making problems with very large
sets of data difficult. The Cray-2s have a very poor method of
interprocessor communication. Having Chen look over Cray's designs,
resulting in the X/MP and Y/MP (and C90s, T90s, etc.), was the key,
resulting in a family of first-class machines.
William Donzelli
william(a)ans.net
26 Apr
26 Apr
9:17 p.m.
At 11:01 AM 25-04-98 -0400, William Donzelli wrote:
I understood that the lack of virtual memory was a conscious design
decision. If you want the best available memory throughput, the last thing
you need to do is add an additional layer of memory translation and even
worse, page faults. If you want to process large datasets either buy more
memory or decompose your problem (neither of which are easy or cheap...)
Huw Davies | e-mail: Huw.Davies(a)latrobe.edu.au
Information Technology Services | Phone: +61 3 9479 1550 Fax: +61 3 9479
1999
La Trobe University | "My Alfa keeps me poor in a monetary
Melbourne Australia 3083 | sense, but rich in so many other ways"
11:46 p.m.
It was a very conscious choice he made, and for many applications, was
never missed. It did, however, limit the applications that could be
tackled by the machines.
The Cray-1s biggest rival at the time, the Cyber 203/205s from CDC, did
have virtual memory, but not quite as we know it. Once a big vector was
started from memory, there was no stopping the machine - I do not think it
would page fault in the middle of a vector load or store.
Crays are short vector machines, keeping them in registers. The Cybers
were long vector machines, with up to 64K elements, and kept them in
memory. Depending on the problem, either machine would beat the other.
William Donzelli
william(a)ans.net
9841
days inactive
9842
days old
3 comments
3 participants
participants (3)
-
H.Davies＠latrobe.edu.au -
peacock＠simconv.com -
william＠ans.net