Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!mcgill-vision!bloom-beacon!snorkelwacker!apple!tahoe!ted From: ted@tahoe.unr.edu (Ted Sarbin) Newsgroups: alt.folklore.computers Subject: Re: Y VAX? [was : TECO on a DEC-System 10] Message-ID: <3479@tahoe.unr.edu> Date: 19 Jan 90 02:02:07 GMT References: <153.UUL1.3#5131@mvac23.UUCP> <457@ns-mx.uiowa.edu> Reply-To: ted@tahoe.unr.edu (Ted Sarbin) Organization: University of Nevada Reno Lines: 141 In article <457@ns-mx.uiowa.edu> jones@pyrite.cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) writes: >From article <153.UUL1.3#5131@mvac23.UUCP>, >by thomas@mvac23.UUCP (Thomas Lapp): >> >> Why did they name it VAX? Well, as my .sig says it: VAX stands for >> >It's not the PDP series, it's the PDP-11 series. The PDP-11 was a 16 >bit machine, the PDP-8 was an unrelated 8 bit machine, the PDP-10 was ^^^ The PDP-8 (and its predecessor the PDP-5) were 12-bit machines! >an unrelated 36 bit machine, and the PDP-15 was an unrelated 18 bit > >That aside, the PDP-11 went through many models. The high-end line of >PDP-11's went as follows (in chronological order with parenthetic remarks >about siblings that play no part in this story): > > PDP-11/20 -- the ancestral machine (OEM'd as the 11/15?) > with no floating point, and no memory mapping. Actually, the 11/20 had a memory mapping option called the KT11A. It was very complex, and made a basically slow machine much slower. Almost no one understood it and very few KT11As were sold. > PDP-11/45 -- faster, with segmented memory address mapping > and a floating point unit that overlapped > floating point computation with scalar > computation. > (the 11/40 was the OEM version of the 11/45?) The PDP-11/40 was NOT an OEM version of the 11/45. The OEM version of the 11/45 was the 11/45. The 11/40 was a microprogrammed, machine which was slower than the 11/45 but much faster than the 11/20. It had a memory mapping system which was a subset of the 11/45's. No second register set and no supervisor mode. It had its own floating point hardware. > (the 11/35 and 11/30 were introduced later?) The 11/35 was the OEM version of the 11/40 > PDP-11/70 -- as I understand it, this was basically a PDP-11/45 > CPU with an expanded segmented memory mapping > architecture and something other than the UNIBUS > to connect the CPU and main memory. It started out to be an 11/45 CPU but by the time the project was done there were stustantial differences. There was a cache betweent he CPU and the high speed bus. There was also a unibus mapping unit so that peripherals which only knew about 18bit addresses could be used with the 22bit addresses of the 11/70. There also were four mass bus adapters which provided a more efficient way for peripherals to access memory than over the unibus(tm). > (the 11/05 was at about the same time?) No, much earlier. The 11/05 replaced the 11/20 for OEMs. > >Now for the folklore: > PDP-11/10 -- announced at the same time as the PDP-11/20. Was an 11/20 like machine with 256 words of ram and 1k words of fusible link rom. Fortunately, none were ever sold. (There was later a low end end-user machine based on the 11/05 called the 11/10.) > > Doug Jones > jones@herky.cs.uiowa.edu Inappropriate text to fool postnews so that this will be posted. .. .. .. .. .. .. Newsgroups: alt.folklore.computers Subject: Re: Y VAX? [was : TECO on a DEC-System 10] Summary: Expires: References: <153.UUL1.3#5131@mvac23.UUCP> <457@ns-mx.uiowa.edu> Sender: Reply-To: ted@tahoe.unr.edu (Ted Sarbin) Followup-To: Distribution: Organization: University of Nevada Reno Keywords: In article <457@ns-mx.uiowa.edu> jones@pyrite.cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) writes: >From article <153.UUL1.3#5131@mvac23.UUCP>, >by thomas@mvac23.UUCP (Thomas Lapp): >> >> Why did they name it VAX? Well, as my .sig says it: VAX stands for >> >It's not the PDP series, it's the PDP-11 series. The PDP-11 was a 16 >bit machine, the PDP-8 was an unrelated 8 bit machine, the PDP-10 was ^^^ The PDP-8 (and its predecessor the PDP-5) was a 12-bit machine! >an unrelated 36 bit machine, and the PDP-15 was an unrelated 18 bit >machine. All were made by DEC, many used compatible hardware at some >level, but there was not one PDP series in any useful sense. > >That aside, the PDP-11 went through many models. The high-end line of >PDP-11's went as follows (in chronological order with parenthetic remarks >about siblings that play no part in this story): > > PDP-11/20 -- the ancestral machine (OEM'd as the 11/15?) > with no floating point, and no memory mapping. Actually, the 11/20 had a memory mapping option called the KT11A. It was very complex, and made a basically slow machine much slower. Almost no one understood it and very few KT11As were sold. > PDP-11/45 -- faster, with segmented memory address mapping > and a floating point unit that overlapped > floating point computation with scalar > computation. > (the 11/40 was the OEM version of the 11/45?) The PDP-11/40 was NOT an OEM version of the 11/45. The OEM version of the 11/45 was the 11/45. The 11/40 was a microprogrammed machine which was slower than the 11/45 but much faster than the 11/20. It had a memory mapping system which was a subset of the 11/45's. No second register set and no supervisor mode. It had its own floating point instruction set which also turned up on the LSI-11. > (the 11/35 and 11/30 were introduced later?) The 11/35 was the OEM version of the 11/40. > PDP-11/70 -- as I understand it, this was basically a PDP-11/45 > CPU with an expanded segmented memory mapping > architecture and something other than the UNIBUS > to connect the CPU and main memory. It started out to be an 11/45 CPU but by the time the project was done there were stustantial differences. There was a cache betweent he CPU and the high speed bus. There was also a unibus mapping unit so that peripherals which only knew about 18bit addresses could be used with the 22bit addresses of the 11/70. There also were four mass bus adapters which provided a more efficient way for peripherals to access memory than over the unibus(tm). > (the 11/05 was at about the same time?) No, much earlier. The 11/05 replaced the 11/20 for OEMs. > >Now for the folklore: > PDP-11/10 -- announced at the same time as the PDP-11/20. Was an 11/20 like machine with 256 words of ram and 1k words of fusible link rom. Fortunately, none were ever sold. (There was later a low-end end-user machine based on the 11/05 called the 11/10.) > > Doug Jones > jones@herky.cs.uiowa.edu