15 Feb
2025
15 Feb
'25
3:13 p.m.
On Sat, 2025-02-15 at 17:19 -0500, Paul Koning via cctalk wrote:
Oh? Certainly by today's standards a drum memory
machine is quite
slow. But in their day they were used for serious work.
For more than 16 years at JPL, we used Univac 1108, later 1110,
1100/40, then 1100/80, then Clearpath 2200 (CMOS ICs instead of bipolar
discrete). In the 1108s, we had eleven core-loads of swap on FH432 (4
millisecond) drums. We ran fifty demand jobs and ten batch jobs, pretty
much 24/7. Univac said it was impossible to do that much, but they
weren't interested in having Tom Lang's modifications or tunings to
EXEC 8 (later OS/1100).