A Second Life for IBM's SSEC?

[quadibloc]

Recently, I ran across this web page:
http://progress-is-fine.blogspot.com/2017/07/the-evolution-of-ibm.html

It included a photo of the SSEC being inspected by people at Harvard;
apparently, after the
SSEC was decommissioned at IBM's downtown headquarters building, and
taken apart, instead of
being put into storage immediately, it was loaned to Harvard in order to
be put into service
there.
In a way, it's surprising that Harvard, instead of, say, Columbia
University was the lucky
recipient, what with the SSEC coming into being because Watson was
miffed at Harvard's Aiken.
Be that as it may, while the photo is credited to a news wire service,
its proximate source
is a children's encyclopedia, so I don't expect to find much additional
detail by pursuing the
trail directly, and Google searches, so far, have also not turned up
much.
But apparently the SSEC had a longer operational life span than
generally realized, having had
a second life that has generally been forgotten.
Maybe not. Maybe they couldn't put it back together and get it working
at Harvard, for all I
know.

But perhaps someone here has encountered this little-known story.

John Savard

[Lawrence D'Oliveiro]

... apparently, after the SSEC was decommissioned at IBM's downtown
headquarters building ...

The Dave Garroway of the computing world ...

[Peter Flass]

Post by Lawrence D'Oliveiro

... apparently, after the SSEC was decommissioned at IBM's downtown
headquarters building ...

The Dave Garroway of the computing world ...

That name’s a blast from the past. I recall as a kid seeing the network do
this new thing called a “remote” from his home in far-away Mt. Kisco. (at
least I think it was him). Times have changed a bit in sixty years, now we
have live TV from Mars.

[Lawrence D'Oliveiro]

Post by Peter Flass

Post by Lawrence D'Oliveiro

... apparently, after the SSEC was decommissioned at IBM's downtown
headquarters building ...

The Dave Garroway of the computing world ...

That name’s a blast from the past.

I think he did a live TV show from a set that was visible to passersby on
the street through large storefront windows, didn’t he? That’s what drew
the parallel to my mind.

Post by Peter Flass
I recall as a kid seeing the network
do this new thing called a “remote” from his home in far-away Mt. Kisco.
(at least I think it was him). Times have changed a bit in sixty years,
now we have live TV from Mars.

I would love to see the Solar System blanketed with robot probes orbiting,
rendezvousing and landing everywhere. Robot rovers, robot aircraft, robot
ships and submarines ... the more countries that get into the act, the
better.

[John Levine]

Post by quadibloc
http://progress-is-fine.blogspot.com/2017/07/the-evolution-of-ibm.html
It included a photo of the SSEC being inspected by people at Harvard;
apparently, after the
SSEC was decommissioned at IBM's downtown headquarters building, and
taken apart, instead of
being put into storage immediately, it was loaned to Harvard in order to
be put into service there.

I don't believe it.

The SSEC was enormous and fragile, and when they removed it from the
IBM office in NYC in 1952 it was obsolete. Many sources note that it
was replaced by an electronic 701 computer. In 1952 Aiken was building
the electronic Mark IV at Harvard so he wouldn't have wanted the SSEC
even if they'd offered it to him.

Either that picture isn't at Harvard, or it's one of the computers
that Aiken built.

[quadibloc]

The control panel of the SSEC is clearly recognizable in the photograph.
And it's obviously not in place in its normal position as part of the
SSEC installation in New York.l

This could, I suppose, be a photo of the console at IBM before
installation at New York, but as the photo and its caption do genuinely
appear in the book claimed as a source, I would presume it to be
authentic - and if I had access to newspaper files, presumably somewhere
there would be a news story about the SSEC being shipped elsewhere to be
put back together and continue its work.
Perhaps, because it was already obsolete, this didn't get far.

John Savard

[John Levine]

Post by quadibloc
The control panel of the SSEC is clearly recognizable in the photograph.
And it's obviously not in place in its normal position as part of the
SSEC installation in New York.l
This could, I suppose, be a photo of the console at IBM before
installation at New York, but as the photo and its caption do genuinely
appear in the book claimed as a source, I would presume it to be
authentic - and if I had access to newspaper files, presumably somewhere
there would be a news story about the SSEC being shipped elsewhere to be
put back together and continue its work.

Looking at some other pictures of the SSEC, I agree that the
arrangement of the console and printers seems different from most SSEC
pictures. But "IBM's Early Computers" flatly says

The SSEC was shut down in July 1952 and dismantled to make room for
the fully electronic IBM 701, the first computer IBM placed in
production.

They were there, they oughta know.

I have written articles for encyclpedias, back when they existed. They
hire subject matter experts to write the articles but the editors are
not experts. I don't know where that picture was but I am sure it was
not at Harvard and it was just an editing error, perhaps by someone
who assumed that the giant semi-mechanical SSEC was somehow related to
the giant semi-mechanical Mark I.

[quadibloc]

It doesn't seem impossible to me that, after the SSEC was dismantled to
make room for the 701, the parts could then have been shipped to another
location to be reassembled, so that the machine could perform additional
useful work. This is why I was interested in finding out if that did
happen, but somehow became shrouded in obscurity.
Since the photograph was of the actual SSEC console, getting it confused
with the Mark I doesn't seem to have been what happened.

John Savard

[John Levine]

Post by quadibloc
It doesn't seem impossible to me that, after the SSEC was dismantled to
make room for the 701, the parts could then have been shipped to another
location to be reassembled, so that the machine could perform additional
useful work. ...

While I am impressed at your belief that encyclopedias are never
wrong, that remains completely implausible. IBM would not have moved
and reassembled an umpteen ton obsolete computer with no historical
record other than three words in a second hand picture caption. And
certainly not to Harvard which was doing other stuff.

There is a long section of "IBM's Early Computers" on the SSEC. It
says they built it in 1946-47 in Endicott, then moved it to the
showroom in Manhattan and dedicated it in Jan 1948. It was enormous,
with 21,400 relays and 12,500 vacuum tubes, It was very slow compared
to electronic computers, 20 multiplications per second. After it was
dismantled in 1952, the 701 that replaced it had 6,000 tubes, 15,000
solid state diodes, and ran 200 times faster.

I also found Charles Bashe, "The SSEC in Historical Perspective" in
the Annals of the History of Computing, 1982, which is similar to
what's in the book. It says

After more than four years of operation, the SSEC fell victim to the
technological advances that marked the advent of IBM’s fast generation
of computers from the production line. It was shut down in July 1952
and was dismantled to begin making room for installation of the IBM
701.

Bashe started working for IBM in 1949 so he was literally there.

When they started the SSEC in 1945 they knew it was a one-off
transitional machine, built using the technology that worked at the
time. By 1952, there were several academic electronic computers, the
UNIVAC had been commercially available since 1951, and the 701 was
under development. Why would anyone want a white elephant that took up
way more space, used way more power, and didn't work very well?

My guess is that picture was taken in Endicott before they moved the
SSEC to NYC, which explains the different position of the parts and
the cramped layout. The SSEC never was anywhere else.

[Lawrence D'Oliveiro]

Post by John Levine
While I am impressed at your belief that encyclopedias are never
wrong ...

I read a book once that was a catalogue of mistakes in the Encyclopedia
Britannica.

Post by John Levine
When they started the SSEC in 1945 they knew it was a one-off
transitional machine, built using the technology that worked at the
time.

It was meant as a publicity project for the idea of a digital electronic
computer, wasn’t it? Hence the location with large windows offering a good
view from the street.

“Peace”.

Or, should that be, “Vootie”.

[John Levine]

Post by Lawrence D'Oliveiro

Post by John Levine
When they started the SSEC in 1945 they knew it was a one-off
transitional machine, built using the technology that worked at the
time.

It was meant as a publicity project for the idea of a digital electronic
computer, wasn’t it? Hence the location with large windows offering a good
view from the street.

That was part of it. Watson was annoyed that Howard Aiken had hogged
all the credit for the ASCC (as IBM called it) or Mark I (as Aiken
called it.) One motivation for the SSEC was to make it clear that
IBM was in the computer business.

By the time the SSEC was turned off, IBM had the 701, a word addressed
binary machine with CRT memory, and shortly the digit addressed 702.
They were designing the and was designing the 650, a slower and
cheaper but more reliable decimal machine with drum memory which was
extremely successful, with almost 2000 delivered by the time they
ended production in 1962.

[Peter Flass]

Post by John Levine
When they started the SSEC in 1945 they knew it was a one-off
transitional machine, built using the technology that worked at the
time. By 1952, there were several academic electronic computers, the
UNIVAC had been commercially available since 1951, and the 701 was
under development. Why would anyone want a white elephant that took up
way more space, used way more power, and didn't work very well?

On the other hand, why would anyone throw out the Atanasoff-Berry machine
just to free up a few square feet of floor space ?

[John Levine]

Post by Peter Flass

Post by John Levine
under development. Why would anyone want a white elephant that took up
way more space, used way more power, and didn't work very well?

On the other hand, why would anyone throw out the Atanasoff-Berry machine
just to free up a few square feet of floor space ?

The SSEC was a wee bit larger than the ABC.

It is a shame they didn't preserve the distinctive looking console,
but most of it was conventional relays and tubes just like the ones
IBM used in all of their other 1940s equipment. They used only designs
they already knew would work so they could build it quickly.

[Scott Lurndal]

Post by Peter Flass

Post by John Levine
When they started the SSEC in 1945 they knew it was a one-off
transitional machine, built using the technology that worked at the
time. By 1952, there were several academic electronic computers, the
UNIVAC had been commercially available since 1951, and the 701 was
under development. Why would anyone want a white elephant that took up
way more space, used way more power, and didn't work very well?

On the other hand, why would anyone throw out the Atanasoff-Berry machine
just to free up a few square feet of floor space ?

The man responsible (whom I knew) for that was later the chairman
of the computer science department and regretted what he had done.

[Lawrence D'Oliveiro]

Post by Peter Flass
On the other hand, why would anyone throw out the Atanasoff-Berry
machine just to free up a few square feet of floor space ?

The man responsible (whom I knew) for that was later the chairman of the
computer science department and regretted what he had done.

And they laugh at us compulsive hoarders ...