On Sat, 4 Jul 2015, Toby Thain wrote: How many individual computers are in that picture? -- David Griffith dave at 661.org A: Because it fouls the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail?

On 07/04/2015 08:07 PM, Brent Hilpert wrote: Does anyone know of similar photos being taken at approximately the same time at LLL or LANL? I assume that there might also be some from NASA Langley... Photos from other heavy-duty number-crunching installations of the time, such as ECMWF would also be interesting. --Chuck

I remember being taken on a tour around Atlas as part of my college course. It must have been about 1967/8. I went back to work at Harwell (RRD next door to Chilton or Rutherford Labs as it was called) when I returned from Germany in 1971. My job was to interface a PDP8/e to a triple axis neutron defractometer. The 8/e only did data reduction and display. I had already built a dedicated hardware system to do the positioning and stepping. The BF3 counter gave a signal proportion to amplitude and the stepping controller the three position co-ordinates. Result a 3D display of the scattered neutron pattern. What was it for? Finding out the structure of crystals. We also had a straight 8 doing data processsing. Rod On 05/07/2015 09:47, Dave G4UGM wrote:

Sigh, CDC... back when Minnesota was a real player in the big iron world. Their original main office used to be hardly more than a stone's throw from where I sit typing this. Being it's the 4th, I'll bet at least one of my neighbors lands a burned-out rocket on top of the building that sits where CDC once sat. On Sat, Jul 4, 2015 at 7:13 PM, Chuck Guzis <cclist at sydex.com> wrote:

You're right, it wasn't the +main+ office.. I was thinking of the one along the south I-94 service road, near Dale St. On Sat, Jul 4, 2015 at 7:56 PM, Chuck Guzis <cclist at sydex.com> wrote:

On 07/04/2015 05:20 PM, simon wrote: Sorry Simon, I thought folks were referring to the BIG iron... --Chuck

On Sat, 7/4/15, Chuck Guzis <cclist at sydex.com> wrote: Definitely multiple systems. There's a VAX in the upper right hand corner that looks downright tiny compared to the rest of the iron. BLS

On 2015-Jul-04, at 6:32 PM, ANDY HOLT wrote: More photos linked from the source page: https://cds.cern.ch/collection/PhotoLab%20Archives%20Images?ln=en The reference names indicate (or imply) the photos are from 1985, things are a little different than what they would have been in 1990. As it happens, 1985 was the year I was working there. Once I had oriented myself I went looking for something unique I recalled, the IBM 3850 mass storage device (tape cartridges in a honeycomb, accessed by a robotic arm). That should be it in the upper left corner of this photo: https://cds.cern.ch/record/1847689 the long narrow thing seen end-on. Note there are much higher-res photos linked at the bottom of the pages, e.g.: https://cds.cern.ch/record/1847689/files/85-1-583.jpg There you can make out all sorts of details, the "Siemens" logo, "7?890" actually, rather than 7890, a MODCOMP mini (lower left), "Control Data" name/logo, an ASCII-art Mona Lisa behind the windows of the 2nd floor mezzanine, a Gandalf terminal switch (mid-left). In the high-res version of the OP photo: https://cds.cern.ch/record/1847692/files/85-1-586.jpg one can see Chuck is quite right: "CYBER 170" is visible on the 'cross' machines. As I recall, the Cray wasn't there that early (1985). I think it was also before they had clustered VAXen. I spent the summer working on VAXes under both BSD4 and VMS. IIRC, the one in the photo was running VMS, the BSD machine was across the road in the DD division (computing services) office building. The room was referred to as "the Cathedral" as I recall. The photos actually don't do it justice, it felt considerably more expansive than the photos suggest. One did feel some degree of awe stepping in there for the first time.

I do not think this is right. The cross shaped machines are not IBM. They look like later model Cyber 170s or very early Cyber 180s. The IBM 3081 is way in the back, behind the tape drives. It looks like a small installation. -- Will

On 07/05/2015 03:01 AM, P Gebhardt wrote: Pierre, I see these as 844s--but for all I know the same drive was sold as the 9756--I doubt that you could tell the difference from the outside. I remember encountering them mostly as replacements for the terrible 821 drives, which had the appearance of an 808, but substantially increased capacity. I don't think any 821s were ever installed at a customer's site, however--I saw the lot of them that we had falling to the CE sledgehammer. They were replaced by banks of 844s--lots of them. I think one site had well over 100 installed, all hooked together with a MAC on at least 2 Cyber 74s. --Chuck

Wow. It isn?t clear how many systems are behind those drives. And I thought CERL (University of Illinois PLATO) was a big setup. When I was there (1975-1978) it had, I believe, about 20 844 drives. Those were all connected to a single 7054 (via extenders of course, which aren?t well documented in the manuals I have seen) and from there to a dual mainframe setup (Cyber 73 and CDC 6500). Both mainframes would do I/O, and on each there were two PPUs involved in the I/O to allow 1:1 interlace on 70-series channels. This setup seriously exercised the drive reservation logic. I ran into a controlware bug there, trying to use the Deadstart function (03uu). It runs a sequence of conventional operations internally in the 7054, but as supplied it did not correctly handle a drive interlock status and would get things wedged up. That wasn?t hard to fix; it gave me my one exposure to the controlware code, which is why I have a copy of it, though unfortunately not a copy of the manual. In theory, a 7054 with extenders can support 64 drives, since the unit number is a 6-bit value. And also, again in theory, you could hook up multiple 7054s if you need more drives than that. So well over a 100, accessible to a single system, is at least theoretically possible. Chuck, what (in this context) is a ?MAC?? paul

Pretty much any of the big mainframes of the day - IBM, Univac, CDC, Burroughs, and so forth - could deal with huge disk farms with hundreds of disks. Often these farms were the largest expense of an installation. -- Will

----- Original Message ----- From: "P Gebhardt" <p.gebhardt at ymail.com> Sent: Thursday, July 09, 2015 9:27 AM ... ----- Reply ----- How many of those rooms full would fit on a 128GB microSD card the size of your pinky's nail? m