Back to On-Line Documents
Back to Ballistic Research Labs, 1961
Comments on
Ballistic Research Labs, 1961
From: Seymour Metz also
alternate e-mail address,
July 08, 2001
-
http://www.ed-thelen.org/comp-hist/BRL61-c.html#CDC-160
Note related machines 160A, 160G and the peripheral processors of the 6600
et al. Also, NCR remarketed the 160; I believe as the NCR 310.
The chart should show core storage as 12 bits/word; I don't recall whether
there was parity.
-
http://www.ed-thelen.org/comp-hist/BRL61-c.html#CDC-1604
Note related machines 1604A and 924. The architecture of the 1604 was later
enhanced as the basis for the 3600 and 3800.
-
http://www.ed-thelen.org/comp-hist/BRL61-h.html#HONEYWELL-800
Note: multiprogramming implemented in hardware.
-
http://www.ed-thelen.org/comp-hist/BRL61-ibm065.html#IBM-650-RAMAC
I've never heard of the 650 being referred to as RAMAC, although the disk
drive (655?) is a modification of the disk (355?) used on the 305 RAMAC.
Note that in order to attach tape or disk drives you needed the optional
magnetic core (60 words of 10 digits plus sign), which came in the same
cabinet as the optional index registers and floating point instructions.
-
http://www.ed-thelen.org/comp-hist/BRL61-ibm7070.html#IBM-7070
http://www.ed-thelen.org/comp-hist/BRL61-ibm7070.html#IBM-7074
The disk information reflects the disk drive originally announce for the
7070, based on the disk for the 305 RAMAC. Once IBM announced the
availability of the 1301 disk drive on the 7070, the older drive ceased to
be a market factor. The 1301 was considerably faster, but I don't recall the
range of seek times.
Notes: the 7070/72/74 could be upgraded to 30,000 words (300,000 digits) of
core storage. The word had a three-valued sign: plus, minus and alpha.
Instructions could only be plus or minus.
-
http://www.ed-thelen.org/comp-hist/BRL61-ibm7070.html#IBM-7090
The 7090, 7094 and 7094 could also have a 1301 disk drive. The 7094 and 7094
had 7 index registers, although you needed to perform a leave multiple tag
mode in order to access them.
-
http://www.ed-thelen.org/comp-hist/BRL61-ibm7070.html#IBM-STRETCH
Note that the model number was 7030. The 7030 could operate on bytes from
1-8 bits in length, even across word boundaries. The disk drive was
reengineered to become the 1301 used on the 7040/44, 7070/72/74, 7080 and
7090/94. See also Harvest.
This machine was probably the most successful failure in history. Although
IBM nominally lost money on it, much of the technology resurfaced on the
7000 series and on S/360.
-
http://www.ed-thelen.org/comp-hist/BRL61-ibm1401.html#IBM-1401
The 1305 disk drive was quickly replaced by the 1311 on the 1401 and the
1301 on the 1410/7010.
-
http://www.ed-thelen.org/comp-hist/BRL61-r.html#RCA-301
Note: like the IBM 1620, the RCA 301 used an add table rather than a decimal
adder. The successor to this machine was the RCA 3301.
-
http://www.ed-thelen.org/comp-hist/BRL61-r.html#RCA-601
Note: this machine is a successor to the RCA Communications Data Processor
(CDP). See also RCA 604.
-
http://www.ed-thelen.org/comp-hist/BRL61-u.html#UNIVAC-490
the 490 and 494 are the basis for a large number of militarized computers,
e.g., 1230, Q-20.
-
http://www.ed-thelen.org/comp-hist/BRL61-u.html#UNIVAC-1103-1103A
the 1103A was, AFAIK, the first computer with an interrupt capability.
-
http://www.ed-thelen.org/comp-hist/BRL61-u3.html#UNIVAC-SOLID-STATE-80/90
Note: The word "biquinary" used for the SS-80/90 does not refer to the same
encoding as the word "biquinary" used for the IBM 650.
-
http://www.ed-thelen.org/comp-hist/BRL61-u4.html#UNIVAC-III
The text claims 24 bits/word, but the instruction format shows a larger
word. I recall a size of 27 bits, which would agree with the instruction
format shown (25 data bits.)
Shmuel (Seymour J.) Metz
|
From Grant Saviers Oct 2001
re: UNIVAC Memory assembly
http://ed-thelen.org/comp-hist/BRL61-u3.html#UNIVAC-I
Deutsches Museum (saw this last Tuesday) has the full CPU with a couple
of memories installed and one spare in their "memories" display area,
both of which were very well curated (btw - the BRL 1961 report has the
U-I arithmetic wrong - it was "excess 3", not BCD!!).
I helped maintain a U-I at Case circa 1963-66.
|
Back to On-Line Documents
Back to Ballistic Research Labs, 1961