ISIS, Osiris, Intellec, PL/M, iRMX, and Intel - 20th century

Most recent revision dated July 30 2021 addition Dec 18 2023 (c)Copyright Herb Johnson 2023.

This document describes the earliest Intel 4-bit and 8-bit development systems and their software. There's also information about subsequent ownership of these products after Intel sold off the rights; and Web links to sites with more information and software.

I developed this document back in 2009, as part of my interests in early Digital Research, a company founded by Dr. Gary Kildall, to produce and sell CP/M. KIldall developed Intel's PL/M and created some of Intel's FORTRAN based cross-assemblers for the 8008 and 8080. But later, I worked on Intel Multibus systems myself (and with Bill Beech), and wrote more Web pages about Intel's Multibus and about ISIS.

So in 2016 I've split this early Intel ISIS page. I'll keep "here" among the Digital Research pages, the early history of Intel's development systems and the subsequent ownership of that property. It was pretty static stuff. But I'll move the links and discussions of relatively recent ISIS and Multibus work, to my "restoration" part of my Web site, with other Multibus work, and updates to other recent work.

In 2016-17, a considerable amount of work was/is done on Intel 8-bit software including ISIS and PL/M. - Herb Johnson

  • Background
  • Before the 8080 and Multibus
  • ISIS operating system
  • ISIS II operating system
  • OSIRIS operating system
  • PL/M
  • ISIS and Multibus ownership
  • iRMX ownership
  • recent PL/M and ISIS Web links on another Web page
  • 2016-17 work by Beech, Main, Ogden et al
  • Older Web links

    Go to my ISIS Web page on my "restoration" Web pages for what I call "21st century" links about more recent work with ISIS and Intel development systems.


    Background

    In the early 1970's, Intel offered a hardware development system which used the 4004 processor on a proprietary set of (non-Multibus) Intel cards to operate a paper-tape and PROM based system for software development, called the "Intellec 4"; for the 4004, and later the "Intellec 4/MOD 40" for the 4040. Details are described in a section below. There was also an 8008 version and an 8080 version, the "Intellec 8" and "Intellec 8 mod 80".

    During 1974-75, a floppy disk based OS from Intel called ISIS was developed and distributed for the 8080 on Intel's MDS-800 Multibus system. A later version called "ISIS-II" supported more advanced development tools and more memory. Intel's version of the programming language PL/M fist supported the 8008, then later the 8080 and subsequent Intel processors; it was also ported to ISIS-II. iRMX, not described here, was a later Intel real-time operating system which ran on Intel processors.

    Dr. Gary Kildall, who founded Digital Research in 1975 to promote CP/M, provided many of Intel's 8008 and 8080 software products including the PL/M compiler and cross-assemblers, which were written in FORTRAN. The earliest CP/M's were sold with an example BIOS for use with a Intel 8080 Multibus system. Intel turned down Gary Kildall's offer of what became CP/M. Further discussion of Gary Kildall and PL/M and CP/M is on my DRI Web pages and other pages linked from it.

    A copy of the Intel "Intellec 8/MOD 80 Operator's Manual",dated June 1974, refers to the "INTELLEC 8I to mean the Intellec 8 with 8080 CPU"; a non-Multibus Intel system. It's likely Kildall got an 8080 upgrade to his Intellec 8 system not long after it became available; this suggests the upgrade was available by mid-1974. More information from this manual is described below. However, information obtained in 2018 suggests that Kildall developed CP/M on a MDS-800 Multibus system.

    I have or had a substantial Multibus card and document collection. But I'd like more information on pre-Multibus Intel Intellec and MDS and other products and info, let me know if you have them.

    Before the 8080 and Multibus

    Intel's Multibus standards describes an 86-pin connector with specific signals and a secondary connector, all on a card of standard size. Multibus was designed as a general bus standard and product; it became an IEEE standard. But earlier Intel products for microprocessor development, were intended primarily to demonstrate designs, or for development only. However, they had an impact on the products developed with those tools (including CP/M); some were used to build actual products. Details about these development tools of the 1970's are difficult to find in the 21st century as they fell out of favor after the Multibus and the 8080 in the mid-1970's. Information here is from published accounts of the period, trade magazines and conference papers; and from surviving Intel manuals.

    Intel's first Intellec-branded products, introduced at least by August 1973, were cards using the 4004 or the the 8008 processor. These were followed by their 4040 and 8080 products, then the Multibus product line. By the mid-1970's, Intel was marketing only their Multibus-based systems and these older non-Multibus products fell into disuse. That said, the Intel 4040 was in use for some time as an OEM embedded microcontroller (that is, built into various non-user-programmable products).

    4004 and 4040

    The Intel MCS-4 User's Manual of Feb 1973, refers to a number of Intel 4004 development products. The SIM4-01 was a complete 4004 chip set with RAM and ROM, and a 72-pin edge connector. The SIM4-02 prototype board had similar features but more RAM and ROM, and two 86-pin connectors. The signals at these edge connectors are simply ROM, RAM and teletype signals, not a "bus". The MCS-4 PROM programming system product, uses a SIM4-01 or -02 and a programmer board (MP7-03) to program 1702 PROMs, as controlled by a Teletype and a ROMmed program. Other Intel products included the MBC4-10 and MBC4-20 interface and control modules. INtel ends the book with examples of how the 4004 or the SIM4 boards were part of non-Intel products such as point-of-sale terminals, data communications, control and business.

    I have an Intel 4040 processor reference book from 1974, which discusses the Intellec 4/40. The 4/40 was used without a disk operating system for development of 4004 and 4040 products. Software was loaded and saved via front panel, punched paper tape, or by ROM; a PROM programmer was one of the peripherals or "modules" available. The 4040-based "Intellec" is described (page 13-8 following) as the "Intellec 4/MOD 40 Microcomputer Development System" and is photographed as a box badged with "Intellec 4/40". Several modules could be added, all connected via an unnamed "bus" not described at the signal level, but clearly not Multibus, possibly like the bus in the Intellec 8. The standard software was the "PROM resident system monitor, the RAM resident macro assembler". No disk-based products or hardware are described.

    I acquired at a hamfest in the 2000's, an Intel MCS Intellec 4/40 system; also an Intel Universal PROM Programmer or UPP.Here's a Web page about those systems. I've started some restoration work on some UPP boards and systems.

    Photos of my Intel "Universal PROM Programmer", show the board set inside. There is a "2716 personality module" with chips from 1980, a UPP-833 or UPP-033 4040 CPU personality module, and a PROM Programmer control module. The boards have 100 pin edge connectors which look like those used with the Intellec 8 and 80 systems described below; and with intellec 4040 system boards.

    8008, 8080

    An Intel ad in August 1973 IEEE Computer "introduces two modular microcomputer development systems called Intellec 4 and Intellec 8", to design "MCS-4 and MCS-8 OEM systems", that being the names of Intel's chip sets for the 4004 and 8008 processors. "These microcomputer sets were first introduced by Intel in November, 1971". (By context "microcomputer" refers to the chip sets, not the Intellec systems.) The Intellec systems were ROM and paper tape based hardware platforms, using code developed on the minicomputers and mainframes of the time. "A standard software package is also included with each Intellec system and a new high level language has been introduced for the Intellec 8". Presumably that's Kildall's PL/M. The ad refers to a systems introduction "at the National Computer Conference and [they] will be available in October." Prices were between $2200 and $2400 per system.

    Intel offered a simple 8008 development system in 1972, using a 8008 processor and memory board "SIM8-01" inserted into a motherboard chassis (possibly called "MP7-02") with lights, switches and a power supply. A 4004 model was also available. I will write more about this system at a later time. My references include discussions of the development of the MCM-70 computer in 1971-72, using this development system, as described by the York University Computer Museum MCM-70 collection. A nice view of a SIM-08 card is at this link.

    A copy of the Intel "Intellec 8/MOD 80 Operator's Manual", 98003-A, dated June 1974, refers to the "INTELLEC 8I to mean the Intellec 8 with 8080 CPU". The Operator's Manual describes a paper-tape based assembly development environment, with text editor, assembler, and ROM-based monitor. Some of the I/O features of the monitor are similar to I/O features of CP/M's BIOS. There's a description of the 100 pins of the Intellec system card set - this is NOT named as "the Intellec bus", merely "a bus", "the bus", "CPU bus", etc. Here's a Web link to a museum page on an Intel intellec 8/80 system.

    ISIS (I) operating system

    The first version of Intel's ISIS floppy disk operating system for the 8080, was apparently available sometime around 1975. It ran on an early 8080-based Intel Multibus system. Until year 2020 or so, only a few fragments of information appeared to be available for it. Later, ISIS V1.1 disks were found and the codes were de-compiled as I'll describe below. - Herb Johnson

    In Dr. Dobb's Journal #21, there was a 1977 article "ISIS: Anatomy of a Real-World Operating System", written by Perry Hutchison of Iowa State University. His article discusses ISIS internals based on his analysis (disassembly) of ISIS "written entirely in PL/M" and as run on an Intel Intellec MDS system. He examined "object code of version 1.2 of 32K ISIS, received Aug 1976 [with] V1.1 of ASM80, V1.3 of EDIT, V1.1 of UPM, and 2.0 of ICE80." He mentions Intel's "MDS-DOS Hardware Reference Manual", that the floppy controller uses ports 78H-7FH, and supports DMA. He lists some ROM MDS monitor routines also. Thanks to Udo Munk for archiving this article with his collection of ISIS documents.

    Hutchison's article references and includes information from an article in Kilobaud Dec 1977 by David Yulke of ADDS, "about ISIS but not by name". Check the Internet Archive or other resources for the December 1977 Kilobaud issue with article "File Structures Simplified". It's not likely Yulke literally had an ISIS OS in hand; the article is about a file system designs leading to one he wrote for a homebrewed 8080 computer.

    Intel's description

    The following is from the "Intel Data Catalog 1977", Ch 13, "Intellec Development Systems". Required hardware for ISIS-I (the only mention of this product in this manual) is an "Intellec Microcomputer Development System, 16K bytes RAM memory, System console, Single Floppy Disk drive." By contrast, in the same Intel document, required hardware for ISIS-II is referenced as an "Intellec Microcomputer Development System, 32K bytes RAM memory, System console, Single Floppy Disk drive". Media is the soft-sectored 8-inch SSSD diskette.

    The floppy controller hardware in the Intellec system was a two-card Multibus FDC channel board and FDC interface board, referred to collectively as the SBC 201 Diskette Controller. A photo of the interface card shows it with a (c) 1975 copyright. (It's not clear that the Intel non-Multibus MDS system ever had a floppy controller card set. I simply don't know for sure. - Herb Johnson).

    The 8080-based "Intellec" is described as the "Intellec Microcomputer Development System" and is photographed as a box badged with "Intellec MDS", and contains Multibus cards. The "2K 8080 Intellec system monitor" provided means to load or save paper tape programs and files for assembly & text editing. The 8080 assembler is identified as "written in PL/M". Floppy disks, disk controllers and disk-based operations are not discussed in the text, but a diskette drive was described graphically as an "option".

    Developer Ken Bergett's recollections

    Thanks to Roger Arrick for making the following interview transscript available on his Web site. Apparently he interviewed Ken Bergett formerly of Intel (See the Q&A below) and then asked Ken to write up his own history of his Intel work focused on ISIS. - Herb Johnson

    In May 2019, I stumbled over this 2017 account of Ken Bergett, of Intel, who apparently developed ISIS and its utilities in 1974. He says he had "already designed and built a DOS at his previous company", probably for a proprietary minicomputer. He also worked on Intel's 4004 and 8008, the latter "was what brought me to Intel" in 1972. "Jim" was named as the other ISIS developer.

    He says ISIS was written "as a single PL/M program" and compiled on Intel's DEC-10 system. But then Ken disassembled the PL/M-produced 8080 code back to assembler, and reduced it, and removed features, from a 12K PL/M program to an 8K assembly language program. The interview goes on to discuss ISIS-II development, to support Intel's ICE-80 software and hardware and to support a native PL/M compiler and assembler.

    Ken Bergett confirms, by the way, the system Gary Kildall developed CP/M upon; and that ISIS and CP/M were independent propositions. Quoting Ken's account:

     "Gary did manage to get one of the early production MDS 800 systems, with 
    the disk controller we built for the ISIS project, trading for something 
    Marketing wanted done, and the development of CP/M proceeded in parallel 
    with our work on ISIS. Gary and I discussed possible solutions for doing 
    file allocation and I/O, and how to manage the disk controller, and such 
    like, but each DOS was developed separately. 

    The "MDS-800" was the first Intel Multibus system; and not the Intellec 8/80 a non-Multibus 8008 or 8080 system. The above, however, claims Kildall got a cobbled-up UART-based floppy disk controller - discussed earlier in the interview as a discarded effort early in ISIS development. Elsewhere, Kildall only describes his failure to build a controller himself, then his successful use of Dr. John Torode's constructed floppy controller.

    Ken's narrative continues with his history at Intel to 1977, working on the 8086 and also producing RMX-80, a multiprogramming environment first supported on the SBC 80/20 board. Roger also has a Q&A interview with Ken about Osiris, another OS for the Intel 8080 which ran on the MDS-800. Ken says he was not aware of Osiris, it may have been produced after he left Intel.

    post 2020 recovery of ISIS V1.1

    An ISIS V1.1 (16K OS) diskette was uncovered in early September 2020 by Jon Hales, a volunteer at the Cambridge Centre for Computing History. Mark Ogden added this disk to his Intel software tools repository. I am not aware of this being elsewhere on the internet.

    The disk image for V1.1 is there. So the file contents are as80, attrib, copy, delete, dir, edit, format, etc., and the disk image.

    Mark Ogden's disassembled/decompiled PL/M-80 sources. Mark has reproduced a number of sources to ISIS V1.1, ISIS-II, and other Intel early software tools; and produced C-program equivalents to run under Linux and Windows. It's masterful work. See my 21st century ISIS Web page for discussion of his Intel recovery work and other's in the current era.

    Other references

    Frode van der Meerer describes the Tandberg 8080 computer which apparently operated with a derivative of ISIS V1.1 OS and tools. Read my discussion notes with Frode for details.

    A typical document of the era for more information is: "Automatic design of multiprocessor microprocessor systems" by Hoo-min D. Doon of MIT; pages 6-14, Annual ACM IEEE Design Automation Conference, published 1977 in the "Proceedings of the Symposium on Design Automation and Microprocessors".

    A paper I've not reviewed yet is: "Intel SIM8-01: A Proto-PC ", Zbigniew Stachniak, York University in Canada; from the "IEEE Annals of the History of Computing", Volume 29, Issue 1 (January 2007), pages 34-48

    ISIS II operating system

    Most of the quotes below are from the "Intel Data Catalog 1977", from Ch 13, "Intellec Development Systems":

    (page 13-29 to 31, quoting) Title: Software - Intel Systems Implementation Supervisor (ISIS-II)

    The ISIS-II operating system resides on the system diskette and supports a broad range of user-oriented design aid software. Total file management and input editing features greatly reduce software development time. the ISIS-II Relocating Macro Assembler, Linker, Object Locator and Library Manager can be loaded from the diskette in seconds. All passes of the assembler can be executeed without the need for user intervention. Object code and listing may be directed to any output device, or stored as diskette files.

    Powerful system console commands are provided in an easy-to-use English context. Monitor mode can be entered by a special prefix to any system command or program call.

    A file is a user-defined collection of information of variable length. ISIS-II also treats each of the standard Intellec system peripherals as files through preassignment of unique file names to each device. In this manner data can be copied from one device to another (i.e. tape reader to tape punch) using the same command required to copy one diskette data file to another. ISIS-II provides automatic implementation of random access files. Each file is identified by a user-chosen name unique on its diskette. Up to 200 files may be stored on each 1/4 million byte diskette."

    ISIS-II SYSTEM COMMANDS
    
    ATTRIB	Assigns spedified attributes to a file such as write protect.
    
    COPY	creates copies of existing diskette files or transfers files from one device to another.
    
    DELETE	Removes a file from the diskette, thererby freeing space for allocation of other files.
    
    DIR	Listes name,s ize and attributes of files from a specified diskette directory
    
    RENAME	Allows diskette files to be crenamed.
    
    FORMAT	Initializes a diskette and allows creation of additional system or data diskettes.
    
    DEBUG	Loads a specified program from a diskette into memory and then transfers control
    	to the Intellec monitor for execution and/or debugging
    
    SUBMIT	Provides the capability to execute a series of ISIS-II commands which have
    	been previously written to a diskette file.

    The ATRIB, DELETE, and DIR commands have the additional capability of operating on several files at once via the wildcard file-naming convention. As an example, the command "DELETE *.OBJ" deletes all files in the diskette directory with the suffix ".obj"..

    --end quotes--

    Programs described in the "Intel Data Catalog 1977" include the 8080 Macro Assembler, Linker, Object locator, PL/M-80 "resident compiler", Text Editor, Library Manager. Also there's a reference to the ICE-80, an in-circuit emulator which includes hardware to plug into the 8080 processor socket of a target system.

    ISIS II versions

    In late 2018, my friends including Mark Ogden discussed the question of available versions of the ISIS II OS. Jon said: "For a quick answer, I looked at Mark Ogden's collection of files extracted from disk images. The OS versions he lists are: 2.2; 3.4; 4.0; 4.1; 4.2; 4.2(W); 4.3 and 4.3(W)." Bill Beech's Web site also has a description, and links to disk-image archives, for various ISIS versions. Bill covers some hardware supported by each version. [Mark in 2021 notes that ISIS 1.1 is now available.- Herb]

    Jon asked about some description of changes and progress across these versions and into iRMX86 or 286. Mark Ogden offered a brief description in correspondence of ISIS-II versions. The document includes a link to Mark Ogden's decompilation of ISIS-II OS's and tools.

    Go to my restoration ISIS web page for links to Mark's and other's archives on ISIS software and recent work done. - Herb

    ISIS III

    Mark Ogden in April 2021 tells me: "As a preliminary note ISIS III appears to use the same floppy disk format as ISIS 4.x, I have more work to do re hard disks, syscalls etc." He says he has ISIS III versions 2.0 and 2.2. It seems to be associated with ISIS system clusters or NDS networking (Intel).

    PL/M

    PL/M-80 is described in a subsequent section (pages 13-33) as an "Intellec Resident Compiler" which "has been substantially enhanced since its introduction in 1973". (That refers to the 8008 version noted in a 1973 Itel ad above.) Another section describes it as a FORTRAN-based cross complier running on minicomputers and mainframes. Intel's early product development tools ran either as cross-products or as paper-tape or ROM-based native products, until both the 8080 and a floppy disk subsystem were developed.

    In a 1975 article written by Kildall, "Microcomputer software design - a checklist" (National Computer Conference 1975, published in conference proceedings), Kildall describes "Intel's language, called PL/M, has been available since mid-1973 through a cross compiler, while National's product, called PL/M+, will be available in mid-1975 ...[and] is designed for the IMP-16 and PACE microcomputers." The article describes PL/M as producing code for the 8008 and 8080 processor. Kildall developed the Intel PL/M product; it's not clear if he developed the National product. A number of other companies provided PL/I-like cross-compilers. Later, Digital Research provided a PL/I subset G native compiler for CP/M-86.

    Another paper is "Systems Languages: Management's Key to Controlled Software Evolution" by Gary Kildall. Proceedings of the 1974 Western Electronics Show and Convention (WESCON), September 1974 ("1974 WESCON Technical Papers", Volume 18, Session 19/2). In this paper, Kildall discusses how PL/M is an example of a high-level language for microprocessor development. He writes: "Several versions of Intel's PL/M compiler have been released since its first introduction in June of 1973" (Ver 1) for the 8008; another 8008 version (Ver 3) in Feb 1974, and the first 8080 version in March 1974. A text version of the article is at this Web link.

    I have much more to say about PL/M, and these and other references, in my Web page on the history of Gary Kildall and Digital Research..

    OSIRIS operating system

    The initial information about OSIRIS comes from Roger Arrick, who some years ago acquired an Intel MDS-800 (MDS-888) system with hard drive and floppies. They boot into Osiris V3.0, but many files in those file systems are ISIS operating files. Some information on Roger's site from Jack M., suggests OSIRIS may have been a short-lived version of ISIS for the MDS-888 before the MDS-225 was produced. (In Greek mythology, Osiris was the brother of Isis.) Further references and Web search in July 2021 appears to have resolved OSIRIS as a modified ISIS OS in the early 1980's to support hard drive operation.

    Roger has recovered the boot-disk files and some hard disk files. See his Web site for more information and details. In July 2021, Roger described an email he recieved:

    "You may have your hands on development platform for targetting the 'BiiN' [under Ada] — The article says the operating system (OSIRIS), applications, development tools, and every other piece of BiiN software was written exclusively in Ada — perhaps the largest non-military use of that programming language." The email also referenced this [1988] DTIC report of interest."

    As I read the DTIC report, it's an Air Force test of an Ada compiler for "validation", running under a BiiN/OS on a BiiN 60 system of 8 processors. No details of the hardware system.

    The Wikipedia article on BiiN describes an Intel/Siemens project, later a spun-off company BiiN, related to iAPX 432 development of fault-tolerant or mission-critical systems in the 1980's. The article claims "the operating system (OSIRIS), applications, development tools, and every other piece of BiiN software was written exclusively in Ada — perhaps the largest non-military use of that programming language." It references bitsavers.org for BiiN documention (a collection of 1988 BiiN documents). It references the BiiN 60 multiprocessor system apparently with either i432 or 690 MX Intel processors.

    My very quick read of the 1988 BiiN documents at bitsavers does not suggest there's information about OSIRIS there. They are about the multiprocessor system and OS of BiiN. This is material about processors and environments outside of Intel's 80XX86 product line. Further Web search on Osiris and BiiN finds Some iapx432 documentation wherein the '432 component release 3.3 was "used in-house at Intel and Siemens for software development for the P7/Gemini/BiiN operating system (code named Osiris)". Whatever the hell that means.

    My inspection of the Osiris diskette files from Roger's site, shows the OSIRIS.LIB file is a short file of symbols and code fragments. The symbols include ASSIGN, QRYPRT CLREF GETCON SETEF TSTEF SYMLIB, ASSGN CLRFLG GETCFG. ISIS.BIN has a text string "(C) Computer Applications" and "OSIRIS V 3.00". ISIS.T0 has embedded strings "Non-OSIRIS System Disk..." and "OSIRIS V 3.0" but other references to ISIS files. a HELLO file has apparent code and text strings: "(C) Computer Applications, 1982" "V1.1 HELLO Program" "Error in CONSOLE function.." "The new assignment for :CI: is ..." and strings about the "Winchester disk". A DATE file has strings "ISIS", "V1.0", strings for month "JANFEBMAR..." and other strings about date as year month day, hour min sec. [This information may have originated from Eric Smith. - Herb]

    So my inspection suggests an OSIRIS development date of 1982 and strong ties to a version of ISIS. I suggested to my Multibus colleagues, that binary comparisons and selected disassembly to a late version of ISIS files, may confirm that relationship. In my opinion a Web search for a company "Computer Applications" will find nothing useful upon those very common words.

    Mark Ogden, also in July 2021, added to Roger's reportage: "From what I can determine the version of OSIRIS you have is written largely in assembly. It has some volume mounting features over and above ISIS but otherwise appears to have similar functionality to ISIS. Perhaps a version was later written in Ada for larger processor configurations."

    Further clues

    Roger identifies a hard drive with his system: "Radio/Switch model 525", attached to a DTC brand SASI (multibus) interface card. I know DTC as a disk controller company; and something about SASI to MFM adapters. So there's likely a DTC SASI to MFM card in that external drive case and one or two MFM hard drives. (Roger confirmed this later.) Sure enough: a Web search "radio/switch inc model 525 hard drive" finds a publication, "Commercial News U.S.A" with the following content (courtesy of Google Books. page 69, quoted in entirity):

    "MEMORY SUBSYSTEM - The "Model 525 Storage System" provides high-speed, high-capacity disk capabilities for Intel Microcomputer Development Systems, and is said to prevent system obsolescence. The storage device is a Winchester disk providing 11 or 22 megabytes of formatted storage. The OSIRIS operating system maintains compatibility with ISIS while incorporating additional features that enhance programmer productivity. This system is priced from $4.995. WRITE: Richard D Tallman, Vice Pres, Mktg, Radio/Switch Inc, Dept CN, 1272 N McDowell Blvd, Petalhuma, California 94952 USA. Telex 449-1365 RASWCH CAL."

    The document is from "U. S. Department of Commerce, International Trade Administration., 1985 - Commercial products" Commercial News USA Annual Directory 1985. My *guess* is that these product-entries are results of press releases to trade journals from company mailings. Details in the DOC document suggest the entry could be a few years older than 1985. Radio/Switch has a US registered trademark and is described as "for radio and telephone apparatus ... first used in 1983"

    I cannot find further information on "Radio/Switch Inc.". I even checked "Computerworld" magazine (Google Books) which carried lots of press releases like this.

    So: the US DOC publication associates "OSIRIS" with a modified ISIS to support a Winchester drive. And Roger's MDS-888 has a non-Intel DTC brand SASI card to operate the drive. So "Radio/Switch Inc" must have added software and ISIS modifications to add the drive to the ISIS file system support (and a date/time function as well). I've asked for more review of the Osiris OS files to confirm this. - Herb Johnson

    ISIS, iRMX and Multibus ownership

    This section describes the origins and history of ownership of Intel's Multibus line and software. In brief, Intel developed the Multibus, an 86-pin bus with a secondary bus connector, in the mid-1970's. In 1999 Intel sold its Multibus I rights to Radisys: see these SEC document extracts for some business details.

    Check my Digital Research Web page for some discussion of PL/M and ISIS. As of early 2009, I do not know who "owns" ISIS from Intel. But a document on the Hitex Ltd. Web site suggests that Intel "jettisoned" and "stop support" of MDS in 1993. Discussion of Multibus and iRMX is below.

    The subsequent owners of Intel Multibus rights and materials was InBus Engineering: as of 2005 the Inbus site said they were "the world's largest supplier of retired Intel Motherboards and systems". But by Dec 2005 U. S. Technologies at ustechnologies.com had a Web page with a letter from Radisys dated about 2002:

    "UST is the exclusive authorized and licensed source for Intel/RadiSys Multibus I and II printed circuit boards in the world. Only UST has the original Multibus I and II test software, test fixtures and engineering updates insuring you the highest quality OEM Multibus product. RadiSys has recently sold their entire Multibus business to UST. We are the world's #1 source for long-term support, including repair, upgrade, and refurbishment services for all Multibus products previously manufactured by Intel or RadiSys."

    I checked the UST Web site in mid-2021. They continue to list "Intel/RadiSys Multibus I and II" among their "Printed Circuit Boards" available as inventory.

    In 2009 I contacted NorthWest Technical and got these comments on Intel property history; "We have been supporting the Intel Multibus platform since the 1980's and in 1989, Northwest Technical was formed. We sell and repair Multibus boards and systems. See our current website for details. Our new www.multibus-international.com website will be going online very shortly. U.S. Technologies did purchase most of the assets and IP from Radisys Corp., who had previously purchased the same from Intel Corp. We consulted for UST before their purchase and still sell and repair boards for them. They mostly do small builds of Multibus I boards for existing customers who still need them." A statement on that Web page says "in October of 2008, we purchased the Multibus division of Zendex corporation" and they acquired inventory and rights. Northwest is apparently doing much the same work when I checked in mid-2021.

    Tenasys - spinoff of Radisys

    In 2008, I originally identified Tenasys as "another company which offers iRMX type operating systems." At that time, they offered a version of iRMX 2.3, and an iRMX for Windows product, with support and current development. Tenasys has a number of PDF'ed iRMX III documents.

    In 2015, while researching further Radisys/Intel documents, I obtained these documents and references.

    The Tenasys company overview Web page says: "A privately held company, TenAsys was founded by a team of software engineers who helped develop the ground-breaking object-based RTOS known as iRMX at Intel in the early 1980s." Thanks to a Wikipedia entry which references this Web page I found further published accounts about the founding of Tenasys.

    Here's a quote from Consulting-Specifying Engineer "magazine" for July 2000, "TenAsys spins off real-time software business from RadiSys", by Staff, 07/01/2000

    quote - TenAsys Corp. (Beaverton, Ore.) and RadiSys Corp. announced May 23 an agreement to transfer development and maintenance of the iRMX and INtime real-time software products to TenAsys. TenAsys will develop, support, and sell the products worldwide. The engineering team includes key developers of the iRMX operating systems at Intel Corp. and the INtime real-time extension to Windows NT at RadiSys Corp. According to TenaSys, this will enable it to exploit the strengths and capabilities of the INtime and iRMX products to increase visibility and market presence, and improve customer relations and support. TenAsys plans to build on an installed base of hundreds of thousands of iRMX and INtime installations worldwide in mission-critical applications. - end quote.

    I checked the tenasys.com Web site in mid-2021. Their "company background" says they have "35 years of pioneering development and knowledge in RTOS and embedded virtualization..." From 2020 that dates back to 1985; the date of Intel's iRMX development. A site search for "irmx" leads to a "legacy" Web page of IRMX III R2.3 for Windows documentation. These 1999 documents have the RadiSys logo on them.

    Other companies

    Other companies also offer iRMX services or related products. Links to these and other Multibus or iRMX vendors are listed below.

    Profimatics GmbH is apparently a company in Germany which offers some iRMX support. Their site has some iRMX manuals in PDF format, as of Dec 2003.

    Hitex Ltd. is/was a UK company which offers software development tools. They have a long history of 8080 based tools and provide some support for use of old Intel tools. On their Web site, I found this document (no longer available there) which discusses the history of use of ISIS, MDS systems and so forth particularly in military applications support. The document says: "Ultimately, the death nell for all the old Intel MDS was Intel’s 1993 decision to jettison its tools arm and stop support for all the old systems." They add "Hitex was one of the leading pioneers of PC-hosted development tools for Intel microprocessors in the early 1980s. Unlike other independent tool manufacturers of the time, Hitex developed PC/MS-DOS interfaces to Intel MDSs and their operating systems. These are still available from Hitex to help in the maintenance of older projects." Hitex's support seems to be through emulators, simulators, and similar tools on MS-DOS based systems which can run the older products. (However, as of 2009 Hitex Ltd. appears to be part of Hitex.)

    2016-17 work by Beech, Main, Ogden et al

  • Recent PL/M and ISIS Web links are on another Web page

    Older Web links

    Go to my ISIS file on my "restoration" Web pages for links on my Web site and on other Web sites, about recent work with ISIS and Intel development systems. Here's a few older Web links.

    Here's an article by Stan Mazor, co-creator of early Intel CPU's. "The History of the Microcomputer - Invention and Evolution", by Stanley Mazor, in the "PROCEEDINGS OF THE IEEE" VOL. 83, NO. 12, PP. 1600-1608, DECEMBER 1995. A copy of the article is here, with illustrations.

    The "bitsavers.org" archive Web site has for many years, a section for Intel MDS disk images. There are mirror sites for that archive site, the home page will list them.

    There WAS a yahoo group, "irmxstuff", for iRMX & INtime Real-time Operating Systems, founded 1998. It was down to several messages a year since 2010. Of course Yahoo ended their Yahoo groups in the late 2010's. One 2010 message apparently from Tenasys staff, suggests support for Intel iRMX PL/M products: "The iRMX For Windows add-on for INtime 3 supports the 386 development tools including PL/M. We continued to use our PL/M and C development for iRMX using iRMX For Windows under INtime 3. - Regards, Rich" [poster "iRMXstuff"]. In 2021 I updated the link; tenasys still has iRMX documentation, as discussed elsewhere on this Web page.


    Contact information:

    Herb Johnson
    New Jersey, USA
    To email @ me, see see my ordering Web page.

    Copyright © 2023 Herb Johnson