Fred Cisin wrote:
I did some FORTRAN on a 7094 at GSFC (SSDC Bldg 26) 40
years ago. They
were using a 360/30 to do I/O for the 7094! Would that make that 360 a
PPU :-?
I did a lot of FORTRAN on a 7094 at UTIAS (University of Toronto
Institute of Aerospace Studies)
about 45 years ago. While I can't be positive, I believe that all of
the input and output was via tape.
IIRC, a 1401 was used to read the standard IBM cards and copy the input
to a tape drive on the
1401. The 7094 would then read the tape after it was transferred to a
drive on the 7094 with
output being to a second tape on the 7094. I think there were probably
a number of additional
tape drives on the 7094 along with a disk drive (washing machine size
and probably only megabytes
in size). Finally, the output tape was transferred back to a drive on
the 1401 and the actual output
became printed pages and punched cards. In my case, the punched cards
were the executable program
from a FORTRAN source program of about 2000 cards. The
executable was
about 300 to 400 cards
which was a substantial reduction since all cards could be read at the
same speed.
If my memory of the system is correct, the Toronto installation had only
tape drives and a small
disk drive as units which allowed input and output. Does anyone else
remember how the 7094
handled input and output? I would guess that if the 360 could handle
the user interface, then the
360 might be a bit slow at computation.
About 42 years ago, I worked for NORTEL on the CDC 3300. That system
allowed many
users to edit their input files (probably 10 to 20 users on terminals)
and submit the files to
the system to run their programs as batch jobs. IIRC, the batch queue
was one user at a
time since that was probably faster considering the limited capability
of a system in 1968.
For the users, all of their files could be held on the computer hard
drives and no cards
were required. Output was printed hard copy.
Jerome Fine