24 Jan
2012
24 Jan
'12
3:07 p.m.
Hi
This link is making the rounds:
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of-memo…
But what type of memory is it? Delay line, CRT tube, Drum?
Bonuspoints if the machine can be identified :)
Cheers,
Pontus.
24 Jan
24 Jan
3:18 p.m.
Almost certainly not a drum (it would need to be many times bigger, I think, to hold the
motor and the heads, given that it's tube-driven). My semi-educated guess would be a
Williams tube (CRT), but I suppose it could be a mercury delay line.
-Dave
On Jan 24, 2012, at 10:07 AM, Pontus Pihlgren wrote:
Hi
This link is making the rounds:
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of-memo…
But what type of memory is it? Delay line, CRT tube, Drum?
Bonuspoints if the machine can be identified :)
Cheers,
Pontus.
4:22 p.m.
>>>
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of-memor
y
But what type of memory is it? Delay line, CRT tube, Drum?
Bonuspoints if the machine can be identified :)
>>>
Note that I have put these thoughts in the order I had them or researched
updates - this makes some of the earlier reasoning incorrect :-)
First, I wouldn't necessarily rely on the description ... the machines with
memory width in "bytes" did not come into use until significantly later than
valves (vacuum tubes) ceased being used for anything other than displays.
The top two cylinders certainly look as if they contain crts, but as far as
I am aware Williams Tubes were a British idea that wasn't used by IBM** (and
I thought that they were about 5" diameter at the face plate whereas these
can be at most 3"). They also don't look like delay lines or anything
motorised (nobody would so easily (!) carry anything of that era that had
rotating memory).
OK, I'm wrong here - the IBM701 and 702 apparently had Williams Tubes
Further web searches lead me to the IBM 706 which was the main memory of the
IBM 701 system. The object looks like one of the 18 units that together
provided 2K 36 bit words of storage (and thus provides 4K bits, not bytes,
of memory). (and, yes, these are 3" diameter at the face plate)
Think that has nailed it.
Andy
8:20 p.m.
On Tue, 24 Jan 2012, Andy Holt wrote:
Further web searches lead me to the IBM 706 which was
the main memory of the
IBM 701 system. The object looks like one of the 18 units that together
provided 2K 36 bit words of storage (and thus provides 4K bits, not bytes,
of memory). (and, yes, these are 3" diameter at the face plate)
An early version of the Whirlwind computer also used Williams Tubas:
"Initially Whirlwind used a modified form of the Williams tube. An
additional flood gun maintained the pattern of dots while a writing gun
was used to alter the pattern. Thirty two such tubes were needed to
provide the 4KBytes of storage that the Whirlwind needed. Given a tube
life of one month and cost $1000 the running cost of the machine was very
high, $1 per bit per month."
http://history-computer.com/ModernComputer/Electronic/Whirlwind.html
Mike Loewen mloewen at cpumagic.scol.pa.us
Old Technology http://sturgeon.css.psu.edu/~mloewen/Oldtech/
25 Jan
25 Jan
7:53 p.m.
On 2012 Jan 24, at 12:20 PM, Mike Loewen wrote:
On Tue, 24 Jan 2012, Andy Holt wrote:
Whirlwind never used Williams-tube storage, at least as I understand
the definitions. I don't believe it is even historically accurate to
say ".. Whirlwind used a modified form of the Williams tube".
Williams-tubes memory (or the preferred "Williams-Kilburn-tubes") and
the first memory system of Whirlwind were both electrostatic CRT-
based storage. The design approach of the Williams-Kilburn-tube-
memory was intended to enable the use of standard or off-the-shelf
CRTs, while the Whirlwind design used specialised holding-beam CRTs.
Causal influences are another matter. The CHM website suggests the
influence was the other way (MIT->Manchester). My understanding was
that they were simply separate developments of a common concept.
--
For reasons explained in a following message, I wouldn't trust that
"history-computer.com" site for much of anything.
Further web searches lead me to the IBM 706 which
was the main
memory of the
IBM 701 system. The object looks like one of the 18 units that
together
provided 2K 36 bit words of storage (and thus provides 4K bits,
not bytes,
of memory). (and, yes, these are 3" diameter at the face plate)
An early version of the Whirlwind computer also used Williams
Tubas:
"Initially Whirlwind used a modified form of the Williams tube. An
additional flood gun maintained the pattern of dots while a writing
gun was used to alter the pattern. Thirty two such tubes were
needed to provide the 4KBytes of storage that the Whirlwind needed.
Given a tube life of one month and cost $1000 the running cost of
the machine was very high, $1 per bit per month."
8:27 p.m.
On Jan 25, 2012, at 2:53 PM, Brent Hilpert wrote:
Whirlwind never used Williams-tube storage, at least
as I understand the definitions. I don't believe it is even historically accurate to
say ".. Whirlwind used a modified form of the Williams tube".
Williams-tubes memory (or the preferred "Williams-Kilburn-tubes") and the first
memory system of Whirlwind were both electrostatic CRT-based storage. The design approach
of the Williams-Kilburn-tube-memory was intended to enable the use of standard or
off-the-shelf CRTs, while the Whirlwind design used specialised holding-beam CRTs.
I tend to think of basically any CRT memory system as "Williams Tubes", which is
obviously inaccurate. I'll have to recalibrate my taxonomy. :-)
- Dave
9:28 p.m.
On Wed, 25 Jan 2012, Brent Hilpert wrote:
On 2012 Jan 24, at 12:20 PM, Mike Loewen wrote:
When in doubt, go to the source:
"Symposium on Williams'-Tube Storage"
http://dome.mit.edu/bitstream/handle/1721.3/39201/MC665_r05_M-1380.pdf
"Initial Testing of a Computer Electrostatic Storage System"
http://dome.mit.edu/bitstream/handle/1721.3/39822/MC665_r11_R-195.pdf
"Storage Tube Program, Project Whirlwind"
http://dome.mit.edu/bitstream/handle/1721.3/38916/MC665_r04_M-102.pdf
Mike Loewen mloewen at cpumagic.scol.pa.us
Old Technology http://sturgeon.css.psu.edu/~mloewen/Oldtech/
An early version of the Whirlwind computer also used Williams Tubes:
"Initially Whirlwind used a modified form of the Williams tube. An
additional flood gun maintained the pattern of dots while a writing gun was
used to alter the pattern. Thirty two such tubes were needed to provide the
4KBytes of storage that the Whirlwind needed. Given a tube life of one
month and cost $1000 the running cost of the machine was very high, $1 per
bit per month."
Whirlwind never used Williams-tube storage, at least as I understand the
definitions. I don't believe it is even historically accurate to say "..
Whirlwind used a modified form of the Williams tube".
Williams-tubes memory (or the preferred "Williams-Kilburn-tubes") and the
first memory system of Whirlwind were both electrostatic CRT-based storage.
The design approach of the Williams-Kilburn-tube-memory was intended to
enable the use of standard or off-the-shelf CRTs, while the Whirlwind design
used specialised holding-beam CRTs.
10:36 p.m.
On 2012 Jan 25, at 1:28 PM, Mike Loewen wrote:
On Wed, 25 Jan 2012, Brent Hilpert wrote:
Thanks for the references, some great reading, but I'm not clear what
your intention or position is. I don't see anything in those papers
to contradict my points, I believe they support them.
The Williams-tube symposium of the first paper was held in Dec 1951,
long (years) after the Whirlwind project had been working at
developing CRT-storage. The paper mentions other projects working on
Williams-tube implementations (SEAC, ORDVAC, Los Alamos,..), but does
not mention Whirlwind in that group. By that time Whirlwind was
struggling with the inadequacies of their CRT memory, my take on that
paper would be that Everett/Whirlwind was following the efforts of
projects using Williams-tube CRT storage, perhaps under consideration
of it being an alternative to their own (non-Williams) CRT storage.
The first paper states: "One of the early proposed advantages of the
Williams'-type storage, that of the relative economy of using
ordinary cathode-ray tubes, seems to be hard-pressed for
confirmation." Along with other discussion in that paper, along with
the paper detailing the Whirlwind design, one can see how different
the Williams-tube and Whirlwind designs were.
On 2012 Jan 24, at 12:20 PM, Mike Loewen wrote:
When in doubt, go to the source:
"Symposium on Williams'-Tube Storage"
http://dome.mit.edu/bitstream/handle/1721.3/39201/MC665_r05_M-1380.pdf
"Initial Testing of a Computer Electrostatic Storage System"
http://dome.mit.edu/bitstream/handle/1721.3/39822/MC665_r11_R-195.pdf
"Storage Tube Program, Project Whirlwind"
http://dome.mit.edu/bitstream/handle/1721.3/38916/MC665_r04_M-102.pdf
An early version of the Whirlwind computer also used Williams
Tubes:
"Initially Whirlwind used a modified form of the Williams tube.
An additional flood gun maintained the pattern of dots while a
writing gun was used to alter the pattern. Thirty two such tubes
were needed to provide the 4KBytes of storage that the Whirlwind
needed. Given a tube life of one month and cost $1000 the running
cost of the machine was very high, $1 per bit per month."
Whirlwind never used Williams-tube storage, at least as I
understand the definitions. I don't believe it is even
historically accurate to say ".. Whirlwind used a modified form of
the Williams tube".
Williams-tubes memory (or the preferred "Williams-Kilburn-tubes")
and the first memory system of Whirlwind were both electrostatic
CRT-based storage. The design approach of the Williams-Kilburn-
tube-memory was intended to enable the use of standard or off-the-
shelf CRTs, while the Whirlwind design used specialised holding-
beam CRTs.
26 Jan
26 Jan
7:44 p.m.
On Wed, 25 Jan 2012, Brent Hilpert wrote:
Thanks for the references, some great reading, but
I'm not clear what your
intention or position is. I don't see anything in those papers to contradict
my points, I believe they support them.
Sorry, I meant to add that it's quite apparent that the
Williams-Kilburn tube wasn't as such used in the Whirlwind, but the M.I.T.
Electrostatic Storage Tube was based on a similar principle, even though
the implementation was very different. You are correct, and I should have
checked more than one source. :-)
Mike Loewen mloewen at cpumagic.scol.pa.us
Old Technology http://sturgeon.css.psu.edu/~mloewen/Oldtech/
8:19 p.m.
On 26 Jan 2012 at 14:44, Mike Loewen wrote:
Sorry, I meant to add that it's quite apparent
that the
Williams-Kilburn tube wasn't as such used in the Whirlwind, but the
M.I.T. Electrostatic Storage Tube was based on a similar principle,
even though the implementation was very different. You are correct,
and I should have checked more than one source. :-)
At least up through the 70's, Williams-tube type storage was still
undergoing development in the form of EBAM (electron beam addressed
memory).
Was EBAM ever commercially deployed? I remember seeing a rack of
them sitting in the hallway at ADL at CDC Arden Hills in the mid-70s
and being told that the project was no longer a hot item.
cf. IEEE "Computer", February 1975: "The Development of an
Experimental Electron Beam-Addressed Memory Module", John Kelly,
SRI. (Available online via IEEE CS subscribers).
Does anyone have such a device in their collection? CHM maybe?
--Chuck
9:05 p.m.
At least up through the 70's, Williams-tube type
storage was still
undergoing development in the form of EBAM (electron beam addressed
memory).
Was EBAM ever commercially deployed? ?I remember seeing a rack of
them sitting in the hallway at ADL at CDC Arden Hills in the mid-70s
and being told that the project was no longer a hot item.
In the mid-1980s, I remember seeing something of the sort in an
industry rag - electron beam semiconductor devices, I think. I have
never seen one, and I am doubting they ever left the lab.
--
Will
10:43 p.m.
On 1/26/12 12:19 PM, Chuck Guzis wrote:
The Development of an
Experimental Electron Beam-Addressed Memory Module
T I - Electron-Beam Addressed Memory Research.
TN - Final rept. Mar 72-Jan 74
AU - Kelley. dohn: Moore, John S. *
CS - Stanford Research Inst Calif*air Force Avionics Lab.,
PO - Aug 74; 200p
CC - 9B: 62A t
DE 'memory devices; *storage tubes: Ad essing: Electron beams:
ID - 'electron beam addressed memories: Ntisdodaf'
SA - Afal-fr-74-176
AP - U7423
RN - Sri-Isu-1778 .
CG - F33615-72-C-t906
PN 9 Af-4150
AB - Work toward the r ization of a practical operational
W P g~h t - Patterson Af b, ohf 0.
PR - pc 65.50/Mf 62.25
Semiconductor devices: Metal oxide semiconductors; Electrostatics
Electron-Beam-Addressed Memory tebam) is described. Consideration
is given to the design of the various components. The
electrostatic microcapacitor or 'mucap' medium i s described and
analyzed. Operation' in the quast-continuous mode is discussed and
its Importance in the attalnment of devlce stability and target
yield is outlined. Measurements on the storage medium are
presented, two pieces of equipment are! descrlbed: the
Sri-sponsored Analog System, designed for making storage target
meaaurments, and an Ebam module, recently delivered to the Usaf
avionics Lsboratory, Design, construction, and performance of
both systems @re discussed. (author)
i
GE's R&D center was also working on a similar device called BEAMOS memory
a bunch of google hits on "GE BEAMOS"
Accession Number : ADA037642
Title : Design, Fabrication, and Evaluation of an Electron Beam Addressable High
Information Density Memory Tube.
Descriptive Note : Final rept. 1 Mar 75-31 Aug 76,
Corporate Author : GENERAL ELECTRIC CORPORATE RESEARCH AND DEVELOPMENT SCHENECTADY N Y
Personal Author(s) : Lemmond,Charles Q. ; Kirkpatrick,Conilee G. ; Possin,George E. ;
Hughes,William C. ; Klotz,T. H. , Jr
Report Date : FEB 1977
Pagination or Media Count : 94
Abstract : BEAMOS tubes, designed and constructed prior to this contract, were tested to
determine their ability to meet various environmental requirements. Although the tubes
were not designed
specifically with these requirements in mind, data were gained during the tests for use in
effecting a re-design of the tube. Two tubes were fabricated embodying the suggested
design improvements.
Other design improvements, brought about by General Electric's inhouse BEAMOS
development program, were also included. The two tubes so constructed were tested and
shown to meet the operational
performance goals. The tubes were delivered to the Army. More specifically, several
improvements were made to the electron optics of the BEAMOS tube. These changes were made
to produce stable
operation of the tube under various environmental requirements.
Descriptors : *ELECTRON BEAMS, *BLOCK ORIENTED RANDOM ACCESS MEMORIES, *ELECTRON OPTICS,
*STORAGE TUBES, TEST AND EVALUATION, MAGNETIC FIELDS, MILITARY REQUIREMENTS, COMPUTERS,
MEMORY DEVICES, HIGH
DENSITY, ENVIRONMENTAL TESTS, ELECTROMAGNETIC SUSCEPTIBILITY, ADDRESSING.
Subject Categories : ELECTRICAL AND ELECTRONIC EQUIPMENT
COMPUTER HARDWARE
Distribution Statement : APPROVED FOR PUBLIC RELEASE
They seem to be density improvements using MOS semiconductor technology of the barrier
grid storage tubes of the
50's like the Radechon
http://bitsavers.org/pdf/rca/storage_tubes
I'll have to dig back through the material from Hobbs we received over the holidays. I
remember seeing the SRI report in there.
11:25 p.m.
On 26 Jan 2012 at 14:43, Al Kossow wrote:
On 1/26/12 12:19 PM, Chuck Guzis wrote:
> The Development of an
> Experimental Electron Beam-Addressed Memory Module
I picked up a copy of the Kelly/SRI paper from the IEEE online site
and saved the PDF (hint on what to do if you're curious).
The Kelly paper basically describes work done for the USAF--the
contract number is F33615-72-C-1906. There's a photo showing the
experimental AF device (Experimental Unit 740101) being driven by a
Macrodata MD-100 memory exerciser.
I wondered if CHM had any real EBAM-related iron.
--Chuck
2:28 a.m.
On 26 Jan 2012 at 17:05, Al Kossow wrote:
On 1/26/12 3:25 PM, Chuck Guzis wrote:
Found the CHM photo of "Alice":
http://www.computerhistory.org/collections/accession/102712000
http://www.computerhistory.org/collections/accession/102711999
Patents:
3,710,352 assigned to Micro-BIt Corporation, MBC was subsequently
purchased by CDC. The EBAM presence in the hallway of ADL is
explained by a May 28,1975 Computerworld article (via Google Books):
"Control Data Corp. is conducting evaluation tests of a prototype of
an Electron Beam Accessed Memory (Ebam) from Micro-Bit Corp. of
Lexington, Mass.
The unit, being tested on a Star 1B computer as a fast paging device,
has a capacity of 1.2M bits distributed in nine parallel channels
with 128K bits each and has an access time of 16 ?sec to any page,
Micro-Bit said.
The memory combines MOS technology with electron-beam access.
The first Ebam product is scheduled to be introduced in mid-'76 and
will have 18 parallel channels, each with a capacity of 4.2M bits for
a total capacity of 75M bits, the firm said."
The remainder of the article is about IBM's work in
superconductivity, although no mention of Josephson is made.
Evidently, as late as 1982, there were still papers being published
on the subject.
I guess the Williams tube survived longer than anyone thinks...
--Chuck
I wondered if CHM had any real EBAM-related
iron.
I'll have to dig more if we do. If we have it, it's probably
miscataloged 
24 Jan
24 Jan
5:05 p.m.
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of-memo…
? ? ? ? ? ? ? ?
But what type of memory is it? Delay line, CRT tube,
Drum?? ? ? ? ? ? ? ?
? ? ?
Bonuspoints if the machine can be identified :)?
My guess would be CRT memory from the 701.
It looks too short to me to be mercury delay line, especially
for 4k, and it definitely doesn't look like a drum. Plus,
I don't recall IBM ever using delay line for main memory.
BLS
8:34 p.m.
-----Original Message-----
From: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org] On Behalf Of Brian L. Stuart
Sent: 24 January 2012 17:05
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Identify: IBM memory
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of
No one ever said it was "main memory", just memory. I agree it looks like
Williams Tube memory and pictures of the 701 here:-
http://www-03.ibm.com/ibm/history/exhibits/701/701_1415bx01.html
show them as pairs of tubes, but the adjustment manual
http://ed-thelen.org/comp-hist/IBM706-WilliamsTubeMemory-RandyNeff.pdf
seems to show more tubes and caps than the picture above...
-memory
? ? ? ? ? ? ? ?
But what type of memory is it? Delay line, CRT tube,
Drum?
? ? ?
Bonuspoints if the machine can be identified :)
My guess would be CRT memory from the 701.
It looks too short to me to be mercury delay line, especially
for 4k, and it definitely doesn't look like a drum. Plus, I
don't recall IBM ever using delay line for main memory.
BLS
5:08 p.m.
On 24 Jan 2012 at 16:07, Pontus Pihlgren wrote:
Hi
This link is making the rounds:
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of
-memory
But what type of memory is it? Delay line, CRT tube, Drum?
The CHM has one of these selfsame units. Williams tube memory:
http://www.computerhistory.org/collections/accession/X1284.96
Cheers,
Chuck
25 Jan
25 Jan
9:20 a.m.
Thanks for all interresting replies. Here is an update:
http://news.cnet.com/8301-17938_105-57364633-1/antique-ibm-memory-box-math-…
Which confirms what you guys figured out. It is a williams type memory
from an IBM 706
/P
On Tue, Jan 24, 2012 at 04:07:04PM +0100, Pontus Pihlgren wrote:
Hi
This link is making the rounds:
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of-memo…
But what type of memory is it? Delay line, CRT tube, Drum?
Bonuspoints if the machine can be identified :)
Cheers,
Pontus.
2:15 p.m.
At 03:20 AM 1/25/2012, Pontus Pihlgren wrote:
Thanks for all interresting replies. Here is an
update:
http://news.cnet.com/8301-17938_105-57364633-1/antique-ibm-memory-box-math-…
Which confirms what you guys figured out. It is a williams type memory
from an IBM 706
Sorry, it's more circular than that. I tweeted her a link to the CHM
page that Chuck posted. Her response to me seemed to indicate it
was the first response she'd had that was correct. So credit goes
to Chuck.
- John
5:47 p.m.
On 25 Jan 2012 at 8:15, John Foust wrote:
Sorry, it's more circular than that. I tweeted
her a link to the CHM
page that Chuck posted. Her response to me seemed to indicate it was
the first response she'd had that was correct. So credit goes to
Chuck.
If so, I'm flattered, but I'm also surprised that I was first. It
took about 30 seconds to Google and find CHM's image.
Kudos really goes to CHM for documenting this stuff.
--Chuck
10:44 p.m.
On 1/24/12 7:07 AM, Pontus Pihlgren wrote:
Hi
This link is making the rounds:
http://gizmodo.com/5878645/this-giant-chunk-of-metal-is-4-kilobytes-of-memo…
But what type of memory is it? Delay line, CRT tube, Drum?
http://bitsavers.org/pdf/ibm/701/pictures/701_Storage_Drawer.jpg
http://www.computerhistory.org/collections/accession/X1284.96
4497
days inactive
4500
days old
21 comments
11 participants
participants (11)
-
aek＠bitsavers.org -
andyh＠andyh-rayleigh.freeserve.co.uk -
blstuart＠bellsouth.net -
cclist＠sydex.com -
dave.g4ugm＠gmail.com -
fraveydank＠gmail.com -
hilpert＠cs.ubc.ca -
jfoust＠threedee.com -
mloewen＠cpumagic.scol.pa.us -
pontus＠Update.UU.SE -
wdonzelli＠gmail.com