Article 134 of comp.sys.cdc: Path: matra.meer.net!news1.best.com!su-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!ais.net!chippy.visi.com!visi.com!news-out.visi.com!not-for-mail From: nad@visi.com (Ned Deily) Newsgroups: comp.arch,comp.sys.cdc Subject: Re: Merced dual-mode vs. Cyber-180 dual-mode (was Re: Merced a VLIW after all?) Date: 05 Jun 1997 01:02:43 CDT Organization: Home Lines: 70 Message-ID: <5n5kq3$rin$1@darla.visi.com> References: <5mn7sf$4ei@tooting.netapp.com> <33932D4A.59E2@nas.nasa.gov> <5mvojb$fku@tooting.netapp.com> <3394993C.2F1C@nas.nasa.gov> <3395b50d.1930156@news.uni-c.dk> NNTP-Posting-Host: thumper.visi.com NNTP-Posting-Date: 05 Jun 1997 01:02:43 CDT X-Newsreader: NN version 6.5.1 (NOV) Xref: matra.meer.net comp.arch:76368 comp.sys.cdc:134 Peter Bjoern: >From my (pretty good) memory, the machines that would run >dual-state (as it was called) were : >810, 815, 825, 830, 835, 840, 845, 850, 855, 860, 865, 870, >875. >The models ending on 0 were cpu-wise roughly equivalent to >the model just above it ending on a 5, but the '0' models >were of a newer and different technology. I don't remember >exactly how they were different. The families, roughly in order of introduction, were: a) 835 b) 815/825, which begat b1) 810/830 c) 845/855 which begat c1) 840/850/860/870 In each family, the differences among the various models were the performance level of the CPU, whether a second CPU was available, and number of PPs available. On all of these systems, one could run one or two operating systems simultaneously: -NOS only -NOS/BE only (although NOS/BE wasn't officially supported on some models) -NOS and NOS/VE -NOS/BE and NOS/VE -NOS/VE only That's a pretty impressive technical feat when you consider how very different the 170 and 180 architectures were, e.g. 60-bit one's complement arithmetic vs 64-bit two's complement. The 865 and 875 were very different beasts. They were essentially successors to the 750/760 using 7000-series module technology, with some changes to allow larger main memory sizes. These were strictly 170-state systems. Their family tree looks something like this, I think: 7600->176 ->175->750/760->865/875 >Other models that would run dual-state were : >990, 995, 960. >The 962, 931, 932 and the Cyber2000 would (will) only run >180-state. The 960 and 962 used the same CPU and memory, I believe, but the 962 only had the DMA-style (I4) PPs, a configuration not supported by 170-state OS's. There was also a 992 and/or a 994 (I'm getting parity errors here), one or both of which were I4-only configured 990s and, hence, only supported by NOS/VE. The various 930 models and the Cyber2000 were the only true 180-state-only systems. Your memory *is* pretty good! (Some things you never totally forget.) 61100 00001 71600 51604 20652 56610 61000 46000 (or is it 21652? Damn!) 01300 00000 00000 00000 -- Ned Deily, nad@visi.com -- [] Article 136 of comp.sys.cdc: Path: matra.meer.net!news1.best.com!uninett.no!news.uni-c.dk!news.uni-c.dk!not-for-mail From: peterb@inet.uni-c.dk (Peter Bjoern) Newsgroups: comp.arch,comp.sys.cdc Subject: Re: Merced dual-mode vs. Cyber-180 dual-mode (was Re: Merced a VLIW after all?) Date: Thu, 05 Jun 1997 10:07:08 GMT Organization: None Lines: 80 Message-ID: <33977b6f.1079134@news.uni-c.dk> References: <5mn7sf$4ei@tooting.netapp.com> <33932D4A.59E2@nas.nasa.gov> <5mvojb$fku@tooting.netapp.com> <3394993C.2F1C@nas.nasa.gov> <3395b50d.1930156@news.uni-c.dk> <5n5kq3$rin$1@darla.visi.com> NNTP-Posting-Host: lgb232.ppp.uni-c.dk Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Newsreader: Forte Agent .99g/32.339 Xref: matra.meer.net comp.arch:76378 comp.sys.cdc:136 nad@visi.com (Ned Deily) wrote: >In each family, the differences among the various models were the >performance level of the CPU, whether a second CPU was available, >and number of PPs available. On all of these systems, one could >run one or two operating systems simultaneously: >-NOS only >-NOS/BE only (although NOS/BE wasn't officially supported on some models) >-NOS and NOS/VE >-NOS/BE and NOS/VE >-NOS/VE only NOS/VE would only run stand-alone from version 1.2.1, prior to that, it had to run dual-state with a 170 partner OS. I think it was because NAM/VE and CDCNET were not ready until the 1.2.1 release, so VE had to use the 170 OS for network access thru VEIAF and INTERCOM. The 960 was one type where NOS/BE was not officially supported. In 1988, I and a person from CDC spent some weeks in Minneapolis working to make NOS/BE run on the 960. NOS/BE needed some modifications, mostly to the deadstart code. >That's a pretty impressive technical feat when you consider how very >different the 170 and 180 architectures were, e.g. 60-bit one's complement >arithmetic vs 64-bit two's complement. Absolutely ! >The 960 and 962 used the same CPU and memory, I believe, >but the 962 only had the DMA-style (I4) PPs, a configuration >not supported by 170-state OS's. There was also a 992 and/or >a 994 (I'm getting parity errors here), one or both of which >were I4-only configured 990s and, hence, only supported >by NOS/VE. Yeah, I forgot about the 992 and 994. >The various 930 models and the Cyber2000 were the only >true 180-state-only systems. > >Your memory *is* pretty good! Well, yes, but it's only 2 month ago we shut down our last Cyber (Cyber2000). Thru the years I have had the great pleasure of having on my site the following models : 73, 173, 175, 750, 810, 815, 825, 830, 960, 990 and 2000. At one time we had 8-10 mainframes running a mixture of NOS, NOS/BE and NOS/VE at the same time. >(Some things you never totally forget.) > > 61100 00001 71600 51604 > 20652 56610 61000 46000 (or is it 21652? Damn!) > 01300 00000 00000 00000 Oh yeah, that would be SB1 1 SX6 =3END LX6 42 SA6 B1 XJ 0 Effectively ending the program by calling the "END" PP, or perhaps END was not really a PP but was processed directly by CPMTR. (okay, I cheated and found my reference card :-) Regards Peter =================================================== Spambait : postmaster@127.0.0.1 ===================================================