Talk on S-100 history, by Herbert Johnson Given April 1? 2015 at Vintage Computer Festival - East 10.0, MARCH/Infoage, Wall NJ Notes made after the talk, with ebellishments after the fact. (C) Copyright Herb Johnson 2015 all rights reserved. LAst update May 6 2015. Introduction I was asked to give a S-100 talk on Saturday afternoon of VCF-East 10.0, when a scheduled speaker lost his voice. The talk was attended by about 50 people and included a number of S-100 community members. I asked John Monahan to give remarks about his recent IEEE-696 work, to complete my discussion of past S-100 activity with his current S-100 design activity. - Herb Johnson * Days before MITS Altair 8080's expensive - hundreds of dollars other competing microprocessors - 6800, National, Fairchild, etc Legacy of 4004, 8008 - hard to build with peripherals expensive, big - hand me down minicomputer, mainframe limited access to computing tools to assemble, edit mid-1970's prices, wages low - cars, tuition, homes limited exposure to computers except accountants, engineers, scientists computing as massive, corporate, private or government people built electronic kits to get best performance, for education, fun - even etched their own PC boards, some layed them out, or bought layouts * Why was the MITS Altair attractive? #1 - price not much above 8080 retail price, $400, later $625 with RAM #2 - front panel allowed programming w/out assembler, Teletype, glass terminal #3 - 100 pin Altair bus, easy to adapt, just 8080 expanded out. #4 - the POSSIBILITY of expansion (but nothing to buy yet) kit - people built kits in those days for audio, radio large board size - 10 X 5, to accomodate TTL based designs, low-density memory cabinet - looked like minicomputer, including front panel MITS reputation - kits for audio, test equipment * How the Altair bus became S-100 The MITS Altair sold fast - more demand than supply TEchies with skills in design, now had a "target", produced their own boards; - next step was to sell those too. Many kitchen-table companies for Altair bus. - next step after that was to produce Altair bus boxes Result: several companies within one year, offering "Altair bus compatible" or "MITS/IMSAI/Cromemco/etc. compatible" this was too complicated a descriptor. besides, it advertized MITS Altair brand. Probably just before or during Atlantic computer show, probably led by Cromemco, was a collective decision to name the "S-100 bus". Cromemco - Cro Memorial Hall, Dazzler color video Morrow - George Morrow, later Compupro TDL - Xitan, early Z80 products North Star - briefly Kentucky Fried Computers, early hard-sectored floppy Kitchen table companies Tarbell - cassette, later floppy, serial * S-100 market exploded IMSAI wanted 100 Altairs, MITS could not deliver - so they designed their own version, also front panel. Became very popular too. Several S-100 companies for systems, many more for individual boards some with unique features. Boards with more memory - 1K, then 4K, then 8K Boards with functions - cassette, video, then floppy LSI boards Beginning of large-scale integrated LSI chips per function. UART already available. 8-bit processor "chip sets" included parallel and serial chips - Intel and Motorola, later Zilog People started "upgrading" - modifying, later replacing boards "Brands" became established: Compupro, Cromemco, IMSAI, Northstar, Ithaca Developed their own complete systems. But also many small "brands" producing inexpensive cards or cards with advanced functions. Time and date, telephone modem, video * CP/M for S-100 Earliest CP/M product was hardware controller board (Digital Systems) with CP/M (Digital Research) and S-100 interface card. Other floppy controllers by MITS, IMSAI were also one-controller board, one-interface board products, with proprietary OS's. Also: hard-sectored single-S100 cards by Northstar, Micropolis, etc. Then CP/M began to become a major 8080/Z80 OS. IMSAI and DRI developed CP/M 1.3. DRI released a generic CP/M 1.4, later 2.0 close followed by 2.2. OTher brands offered CP/M on their hardware. production of Western Digital 1771-series floppy controller chips and data-seperator designs, resulted in floppy-controller products on S-100 by tarbell et al. These would run single density in a uniform "IBM 3270" low-level format; CP/M established standard sector, directory format. Later, double density operation with advanced chipsets. Intel floppy controller also. * other brands too A brand example: the Apple II. A sales "driver" was Supercalc, first available on the Apple II. Businesses, accountants would buy the Apple II just for the spreadsheet! Tandy was another early brand, with the TRS-80. SS-50 (6800 and later 6809 Motorola CPUs) was popular and comparable to the S-100 "market". various other niches: Digital Group a typical one. * IEEE-696 and multi-function cards, market by brand IEEE standard persued by George Morrow, Bill Godbout, others, in late 1970's. Early version followed by Compupro; standard not finalized unil 1983 as 696. Standard withdrawn in ?late 1990's? Multiple brands of chipsets for floppy, serial, parallel, video, hard drive higher-density memory: 16K DRAM, ROM, SRAM, continued to increase capacity Hard drives expensive, controllers large and complex. SMD 8 inch and larger; - later SCSI and MFM 5.25-inch; smaller controllers. Very non-standard until - Western Digital chipsets, and much later IDE standard. Result? better inter-operability among different brands. But S-100 market now established "by brand" plus various add-on cards. Heath/Zenith produces Z-100: motherboard with dual 8085, 8088 CPUs plus memory, I/O, CGA video; S-100 for floppy, more RAM, hard disk control, more serial, etc. Compupro 8/16 system. Backplane motherboard, CPU 85/88 similar to Z-100, mutifunction I/O card, RAM cards, floppy controller card, later video cards. Cromemco Systems with Z80 Hundreds of brands of S-100 products by early 1980's. * late 1970's gaming computer market Companies like Commodore, Atari produced microprocessor systems which operated video games using TV screens. Fairly simple video designs, complicated audio chipsets, gaming controls, ROM carts (and audio data cassettes). These systems cost only a few hundred dollars - much cheaper than S-100 systems at a thousand dollars. Also, many more of these were produced, and distributed by "department stores", computer stores, etc. Common remark in retrospect: "I wanted a S-100 system but could not afford it. So I got a (popular computer or gaming system) instead." There was a "bubble" of video-game systems which collasped in the mid-1980's. They became super cheap, and available at thrift stores and flea markets. So: almost anyone in America who was a kid in the 80's and 90's, was exposed to these systems, or owned them. * 1980 personal computer market Architectures like S-100, SS-50, Apple II, many single-board computers. Operating systems: CP/M, Flex, Apple DOS, proprietary DOSes General software: word processing, telecomm, database, spreadsheet Mass storage: floppy, 5.25-inch hard drives Displays: mostly text, but also graphics and color (CGA class) peripherals: dot matrix and full-character printers, terminals, monitors Price: $2000 a typical price-point, often twice that Remember: the MITS Altair began in 1975; other microcomputers just before or after. hundreds of computer stores, mostly mom-pop (except Computerland via IMSAI), some per-brand. Thousands of salespersons, tech repairpersons, programmers, system integrators. Many more private individuals doing the same thing for themselves and a few others. This established the personal computing market and collection of experienced persons, that later "big brands" tapped into, to service and sell THEIR brand of computer. * 1981: the IBM PC The IBM PC made the personal computer legitimate for big-corporate America (Fortune 500). IBM tapped the experienced people and small companies and they became "converted to IBM". MS-DOS itself, was a functional "copy" of CP/M (but not a code copy). While IBM "took all the oxygen in the room" in 1981, 82; they lacked software. It had to be converted from 8080 to 8088; from CP/M to MS-DOS. Also: IBM PC's were not cheap; many people kept their CP/M systems and continued to use them. And non-PC's continued to sell *new* for some years. * S-100 in decline By the late 1980's, the advantges of S-100 became disadvantages. Large boards not needed, small ISA boards sufficient, the minimal system was "sufficient". Big boxes were out of vogue. IBM PC's were being "cloned", costs went down somewhat. IBM PC's offered all the new software, and most of the old. The IBM "brand" and brand support was compelling. The IBM PC became a "standard", 100% compatibility became a requirement. But IEEE-696 systems were STILL being designed, produced and sold into the 1990's. Advanced video production systems by Cromemco. Multiuser, multiCPU systems under TurboDOS, many brands. High speed 286, 386 systems by Cromemco, Compupro. Specialized, industrial-class systems, many brands. Eventually, the "clone PC's" became relatively cheap and plentiful. Even older PC's became useable on the used market. Except for specialized markets, IEEE-696 systems fell out of use and production. S-100 systems (by contrast against video-game systems) almost never appeared in thrift stores - too "industrial". If they weren't still in use (many were), they were often scrapped. Or they made it to hamfests, electronic surplus stores. S-100 in the 21st century Of course, vintage computing interests in restoration, operation. Hobby design efforts to make new S-100 cards. Restoration efforts to remake old (mostly MITS like) S-100 cards. Ready market of original S-100 cards from old hoards in the Web marketplace. Comments by John Monahan of s100computers.com John spoke briefly from the audience about his IEEE-696 activities. He's working on an 80486 CPU board which takes full advantage of the expanded "S-100" bus, as does previous boards he's designed. His experience suggests 10MHz to 12MHz is about as fast as the bus will operate. He noted how 32-bit memory management can be resolved with some "over the top" CPU-to-memory board cross-connectors. He made the larger point that a number of designers in recent years, have developed small runs of S-100 cards for hobby use, with 1990's and later computing chips and technology.