30 Jan
2014
30 Jan
'14
9:49 a.m.
On Tuesday (01/28/2014 at 09:44PM -0800), Chuck Guzis wrote:
On 01/28/2014 05:38 PM, Chris Elmquist wrote:
It was pretty much everything we could do (technically, practically and
financially) just to keep the nitrogen liquid :-) So, I guess in that
sense it was just cold enough to stay liquid and not much more.
Well, it was a totally closed-loop system with the
gas taken off the top
of the cryostat, fed back through a cryogenerator to make liquid again and
unless there was a problem, the CPU(s) stayed entirely immersed in liquid.
I remember the "wet nitrogen" thing was in response to a question
I'd asked him about "At what temperature do you keep the liquid
nitrogen?", so I suppose that was his way of saying "not too cold".
Who did the IC fab for the system? Fairchild?
Honeywell. We then also engaged a second source in the valley called
PSC but that was pretty near the end so I don't think they built too
many parts before they were no longer needed.
I worked in the group doing the chip design and specifically on a test
system that connected to the BIST (now known as JTAG). We built a tester
out of a Zenith Z150 PC, a couple GPIB instruments and a little custom
hardware that we could deploy to the silicon vendor to use for their
final test process-- so, we were making plans to make it easy for them
to send us good stuff.
Did Neil ever get to build his Super-X-by-God machine?
I recall
that when I expressed some concern over his hard partition of 16
Kwords for the operating system, he answered, something to the
effect of "If you can't fit an operating system into 16K, you don't
belong in this business."
Ya. You often had to work on NRL if you had a position that was different
than his. It was pretty difficult to ever revisit any decisions already
made I do remember that.
There was a lot of designing and thinking about next generation stuff
but we never built anything beyond the ETA10. I do remember discussions
about "intelligent memory", where transputers were going to be distributed
throughout and an integral part of the memory subsystem but I don't know
that I ever heard exactly what they were going to do. NRL was all over
transputers then and had a bunch of them setup on his desk and hooked to
his AtariST.
Old times...
Indeed. Pretty formative years for me. We try to keep that history
and memories of that experience alive around here. The annual reunion
of the ETA shutdown doesn't happen anymore although this year will be
the 25th anniversary so maybe something will come together.
I still have lunch on a regular basis with many of the guys I worked
with then. A lot of us were hams and so we have that in common too.
I can still say I have never worked for a better place, with a better
bunch of people than I did those years from '83 to '89.
Chris
--
Chris Elmquist