Last Update: 21-Oct-2001
Welcome to my VAX pages! I accidentally got into VAX collecting when I bought two at a sealed bid auction for $26. One was in a DEC BA123 cabinet and the other was in a third-party BA123-like cabinet (Sigma Information Systems). Because I had this old computer, and I'm fond of old computers, I decided to find out what I could about it (lots) and then discovered that they are a lot of fun to play with.
NOTE: All images on this site are COPYRIGHT © 2000, 2001 by Charles McManis, you may not use them without written permission from me! (Especially EBAY Auctions!) You may put a link to this site with a note that your system is "like" one of mine but you cannot use the images. Thank you.
There are several subsections to this area, if you've been to other portions of the site they should be familiar:
I'm a big fan of NetBSD running on these VAXen. As I develop things for them I will put links to them here. Currently I've got two interesting links: [Sorry these are out of date, I'll update them when I get 1.5.2 running on my VAXen soon! -- Chuck]
- netbsd.mv2.gz -- This is a 1.5 kernel build with just KA630/KA650 support for small memory systems.
- MVII -- This is the configuration file I used to configure the above kernel.
VAX stands for Virtual Address eXtension, and it is basically the PDP-11 architecture (16 bits) extended to 32 bits with support for paged virtual memory and twice as many registers. It was introduced in 1977 in the form of the VAX 11/780. With a clock rate that allowed it to execute one million instructions per second, and a price point that allowed many universities and other institutions to buy one, its popularity and accessibility helped the VAX become the "gold standard" to measure against. Based on what a VAX could do, The first "chip" version of the VAX architecture was the MicroVAX 1. It was only 1/3 the performance of the 11/780 and only slightly faster than the 11/23. Its CPU Module ID is KD32, that would make it a PDP-11 CPU which were generally KDx11 modules. It used the same bus as the PDP-11 (called the LSI-11 bus) but that was later rechristened the Q-Bus. With the introduction of the MicroVAX II in 1985 at 90% the speed of an 11/780 the age of the Q-Bus based VAX officially began in earnest.
Carnegie-Mellon University described the "ideal" computer in a paper describing the future computing environment of choice [does anyone have the exact title? I've not found it yet --Chuck]:
This became known as the "3M" machine specification and was the definition of a "workstation" for a long time. Of course the use of a Digital Equipment Corporation machine as the standard by which to measure performance was met with not a little bit of controversy in the computer world.
The center of the controversy was that the term "MIPS" with originally stood for "Millions of Instructions Per Second" became more commonly known as the "Meaningless Indicator of Processor Speed." [I credit this re-interpretation to John Mashey who authored some very good analysis of instruction architecture. --Chuck] The controversy centered around the difficulty of comparing dissimilar architectures using something as mundane as the number of instructions it could crank through in one second. In particular, the RISC chips that were introduced in the mid-80's could claim "10 MIPS" but that did not mean they got 10x the number of things done as a VAX! .
However, a couple of things that happened that certainly affected my life were that DARPA (the Defense Advanced Research Projects Agency) contracted with the Computer Sciences Research Group to port a version of a (at the time) new and experimental operating system called UNIX to the VAX architecture. As I understand it CSRG had already ported a version of their UNIX to the PDP-11 and it was logically just an extension of that work to make it run on a VAX. Further, it gave DARPA access to an operating system that was "free" and unencumbered, which VMS most certainly was not.
My first encounter with this OS was in college when my wife-to-be was working on a project to port it to a Data General Eclipse machine, but my first meaningful encounter was when I went to work at Sun Microsystems in the mid-80's. I joined Sun after proposing that Intel should build a system that would later be called a "Workstation" out of the 80386 and a graphics chip I had been working on. Intel wasn't convinced that there was much of a market for workstations beyond MCAD and ECAD type design engineers and while they were aggressively (for them anyway, in a chip recession no less) pursuing being in the "systems" business, they were ignoring this segment. I on the other hand had a wife who was working at Xerox in Palo Alto and she was using a "Dandelion" and "XDE" to do code development and it was waaay cooler than 80 x 24 terminals any day. Of course Xerox (corporate) didn't have a clue about workstations either, but Sun did. I joined in 1986 and never looked back.
Good question! Well as I mentioned when I started, I ended up with two MicroVAX II systems for $26. The MicroVAX II was in fact DEC's response to the CMU "3-M" machine spec, and with a graphics interface you could nearly build a 3M machine (0.9 MIPS, 1MW (4 Mb) memory, and ~1MPixel (1024x768)). [ Sun on the other hand had a display that was 1152 x 864 which came much closer to actually being 1Mpixel --Chuck]
And while I worked at Sun, DEC was one of the "bad guys". However, I had a good friend who worked at DEC and had paid a good chunk of my college tuition by programming an 11/55 and a variety of KL-10s (USC-ECL) The lynch pin was that the NetBSD Project had started an effort to port NetBSD to the VAX (how's that for coming full circle!) and I thought it would be fun to help out that effort in my spare time.
Back to Chuck's Turn of the Century Computers Page
[Content last updated 10/21/01]
Copyright (c) 2000, Chuck McManis, All Rights Reserved