To: Distribution From: David K. Kahaner ONR Asia [kahaner@xroads.cc.u-tokyo.ac.jp] Re: TRON (The Real Time Operating System Nucleus) 4 March 1991 ABSTRACT. TRON is a large and long term project to develop a computer Operating System (OS) along with a global man-machine interface that can work with many other computer operating systems. Its main focus is to provide an environment for a very large number of small distributed computers to cooperate in real time. The project is supported entirely by industry, mostly Japanese. Western academic computer scientists should make themselves much more knowledgeable about the details. SUMMARY. TRON (The Real Time Operating System Nucleus) is a complex and controversial subject, especially in the United States. It is an ambitious project which is attempting to develop an Operating System (OS) specification that will be coupled to a global man-machine interface that can work with other, different computer operating systems, such as those of IBM, SUN, APPLE, and DEC. Its main focus is to provide an environment for a very large number of small distributed computers to cooperate in real time. The project is supported entirely by industry, mostly Japanese but a few Western. The originator and principle investigator of the project is Prof Ken Sakamura Department of Information Science Faculty of Science University of Tokyo 7-3-1 Hongo, Bunkyo-ku Tokyo 113 Japan Tel: +81-3-3812-2111 x 4094, Fax: +81-3-3779-5753 Email: SAKAMURA@TANSEI.CC.U-TOKYO.AC.JP Professor Sakamura created the TRON concept in 1984 and has been working energetically to promote industry's acceptance of it since then. It could have an important impact, when completed and if accepted, in many areas of information technology, such as advanced numerical controls, robotics, task interfaced plant operations, and simultaneous inter-task communications as well as in many consumer applications. Sakamura claims that all the funding for the project has come from industrial sources through the TRON Association. Membership now numbers almost 150 and includes essentially all major Japanese electronics companies, as well as construction companies, software houses, etc. (A complete list is given at the end of this report.) There are also a few US-Japan subsidiaries such as IBM Japan, and Apple Japan and a very small number of US companies such as Tandem Computer and Motorola. A great deal has been written about TRON. The April 1987 issue of IEEE Micro was entirely devoted to this topic and Sakamura received an IEEE award for best paper of the year because of his work. TRON was also the subject of a short chapter written by Professor Michael Harrison in a JTEC report "Advanced Computing in Japan", Oct 1990. Both these references also contain citations to many other research papers. TRON has been directly involved in basic trade friction between the US and Japan and this has obscured some of its major technical aspects. For example more than 50 percent of Harrison's report was concerned with trade concerns. Even so, in the West, TRON is not well known, and details about the project are much better known in the industrial community than in the basic research community. The purpose of this report is not to repeat material that has already been covered in other sources or to deal with contentious trade problems in any way, but rather to point out to research scientists that there are TRON concepts that have wide implication in many other aspects of computing science. Why is TRON important? It is not necessary to master the details of the various TRON architecture specifications to grasp its potential significance. Sakamura is visualizing a world even more computerized than today's. In such a world one's everyday life will be influenced by computers that adapt the local environment, provide basic functions for communication, transportation, and control, as well as interaction between other heterogeneous computer environments. What are the key characteristics associated with such a computerized world? Clearly, these include distributed processing, toleration of faults, rapid response (real time) to multi media inputs and outputs (voice, image, text, and others), a very human and friendly interface, and adaptation to a constantly changing universe in which the network of computers will be undergoing constant reconfiguration as nodes are added, removed or replaced by different systems. These are some of the same issues confronting researchers in scientific parallel and distributed computing. (Of course, there is a difference between the needs of high performance distributed numerical computation, and real time distributed computation, but this difference is primarily related to task switching speed rather than fundamental organization.) But one major difference between the TRON and other distributed computing models is that TRON is driven by large numbers of small sensor-sensitive computers, the kind that are most likely to make their appearance in inexpensive consumer devices, light pens, touch screens, television, etc. The important part of this is that in Japan consumer electronics is fueling the computer industry rather than the other way around. Consumer product specialists such as Panasonic, Sony, Seiko, Ricoh, Kyocera, Oki, etc., are deeply involved in computer developments. When we realize that in the US one of the most successful parallel computers, Intel's iPSC (hypercube), got its first boost by being able to make use of large numbers of inexpensive, off the shelf processors that were used in PSs, it is possible to imagine how the transfer from consumer technology to computer technology occurs here. One very good (and well known) example is the TRON House. This is an extremely modern residence of almost 400 square meters situated in the center of Roppongi, one of the most expensive sections of residential Tokyo. The general appearance of TRON house is similar to what one would find in the pages of Architectural Digest-modern, open, functional. The building connects about 1,000 computers that are all linked together to perform a dazzling variety of automated functions controlling lighting, heating, cooling, domestic hot water use, ironing, cleaning, personal hygiene equipment, entertainment, ticketing, etc. Currently, a family is actually living in the house at the same time that experiments are being conducted using the computers. (For example, what happens if some of the ystems break down.) Plans are to open the residence in April 1991 to public inspection. The project is supported by nineteen Japanese companies including NTT, Nippon Homes, Mitsubishi Electric, Toto, Yamaha, etc. The construction industry is Japan's largest, and since 1987 new- housing starts have been averaging about 1.7million per year. The average Japanese new house is about 136square meters (about 1400square feet). You can buy a two floor California style house of that size which sits cheek by jowl against its neighbor and about one hour's train ride from Tokyo for about $700,000 US. Thus TRON house represents a substantial investment. Sakamura plans to extend the concept to a TRON office building, and even to an TRON city with "billions" of cooperating computers. The construction companies see in TRON a future of intelligent buildings, intelligent communities, and intelligent cities, in which non-obtrusive computers will control various functions in environmental control, security, communication, health, amenity, recreation, and transportation systems. A pilot intelligent building is to start construction in 1991. In addition, there is a ten-year project to build a computer city incorporating TRON concepts. Land has already been set aside in Chiba Prefecture and this project is being carried forward by over 40 corporations. Certainly, the research necessary to make such a distributed computing project work must have some relevance to other distributed/cooperative computing projects. The one Sakamura lab that I visited was a feast of gadgets and experiments. There seemed to be no end of equipment, and high-end workstations were packed almost wall to wall. (His lab space occupies virtually an entire floor in the University of Tokyo's Faculty of Science building, although I was told that he is also moving into another larger facility in a different part of the city.) Sakamura has designed an ergonomic keyboard for a TRON-based Gmicro workstation that he demonstrated to me. It was hooked up to a standard Sony video camera and the operator could open a window that displayed the camera's output, which could be processed in real time. The workstation is definitely multimedia capable. There are also wireless electronic pencils (functioning like a mouse) and wireless erasers. He is designing "intelligent" glasses that will sense distance from the screen and adjust text image size accordingly. A video center also lets Sakamura and his students experiment with interaction to and from CDs. The TRON project is now so large that it has branched; BTRON, ITRON, CTRON architectures are associated with business, industry (robotics), and communication respectively. For example the target application classes for CTRON systems include switching and communication, information processing, and workstation applications as central file servers in wide area networks or as hosts in large databases. The CTRON specification defines mechanisms to meet the special needs of hard realtime constraints, high reliability and high performance. Each of the subprojects has ambitions goals. For example BTRON workstations use a 16bit character code to allow them to support many different languages. The BTRON specs make it likely that the system will have real time response capability even if equipped with advanced man-machine- interfaces. The design also incorporates psychological techniques such as requiring users to confirm that they want to process sequences that will take a long time, and the updating of only those portions of the display requiring user attention. BTRON defines a common data format for graphical data, and the BTRON OS has utilities for editing and displaying such data as well as hypertext functions and the ability to link documents in network fashion. There are a number of committees and special interest groups. For example, the automotive committee is studying ways to use TRON in navigation, and safety. There is a TRON computer education research group, consumer electronics research group, intelligent house research group, physically handicapped needs group, etc. The seventh annual TRON Project Symposium was held (in English) in December 1990, coincidentally with the TRON Show where products are displayed and demonstrated. Each Symposium Proceedings is published by Springer-Verlag ("TRON Project 1990", Ken Sakamura (ed), ISBN 0-387-70066-8 Springer-Verlag New York Berlin Heidelberg Tokyo). A list of the titles of the 1990 Symposium is attached to this report. Sakamura has repeatedly emphasized that he is involved in basic research and wants an open system in which creative ideas are shared. For example, in the latest Symposium, he gave a paper on programmable interface design for highly functional distributed systems. The idea here is that cooperation among elements in a large distributed system is only possible if some standard interfaces are provided on all the communication paths in the network. Such interfaces need to be defined between application programs, data formats, network protocol, printer control codes, human/machine interfaces, etc. But if the standard is fixed this will tend to stifle incorporation of new computer technology. Alternatively, if the standard is updated via dated versions, then version inconsistency will soon be a problem. Sakamura proposes to deal with this in the following way. A system with which communication is made can be programmed, and interface specifications can be changed whenever needed. When communication takes place between systems, first the interface specifications on both sides are compared and if necessary, the side requiring higher-level specifications sends a program to the other side, establishing the necessary communications. For example, consider a system in which an ITRON-controlled air conditioner is operated by a BTRON computer. If the air conditioner sends to the BTRON machine a dialog window program for its control, interaction with people can then be left up to the BTRON machine, while the ITRON side need only receive instructions that have been determined based on a standard interface. TRON specifications are published and available for everyone to examine. Members of the TRON Association have access to some additional information, but special provisions are in place to provide academic researchers with full details. Given the involvement of Japanese industry in the TRON project it is not surprising that much of the research is being done at corporate laboratories, but it is somewhat disappointing that there has been so little involvement from the West, and essentially none from the academic community. At the 1990 TRON Symposium a few Western scientists did give papers, but with the lonely exception of Professor James Mooney from West Virginia Univ, their affiliations were entirely industrial. And while Mooney does make recommendations and assessments about CTRON, his paper is mostly concerned with general issues of software portability rather than about detailed TRON research. On the other hand, papers were presented by researchers from NTT, Toshiba, Hitachi, Matsushita, UTokyo, NEC, Mitsubishi, Oki, Fujitsu, and others in Japan. Topics range from the very general such as Sakamura's above, to detailed implementations on Unix, PC and other systems, floating point, graphics, etc. Sakamura admits that there may have been some misunderstanding related to trade problems and that he personally does not know too many Western computer scientists. He explained that while many Japanese scientists like to study Western papers to then generalize and extend them, his approach has been to try and develop his ideas entirely independently. The fact that his support is only from industry may also have diluted interest from the basic research community in the West. However, now some Western companies have expressed their interest by joining the TRON Association. The latest is Tandem computers, who are hoping to expand their knowledge of how to make large computer systems more fault tolerant. (A TRON-specification extended bus, TOXBUS, has been developed specifically to improve the performance of VME or Multibus for tightly coupled high performance systems, as well as fault tolerant systems.) TRON has standard functions such as interrupt, exception handling, and memory control functions, in addition to such basic capabilities as input/output, file management, and debugging. U.S. operating systems with some similarity to TRON are RMX-86, MTOS-68K, VRTX/68000. Users can write in C, C++, Fortran, Pascal, and Forth as well as TRON-specific language (TULS). (Green Hills Software in the US provides compilers for TRON architectures.) About half dozen TRON-based microprocessors are already commercially available, as are a number of ITRON products. Tables describing these are given below. One characteristic of TRON CPUs is the use of a large linear (non- segmented) address space. The early design was for a 32bit CPU with expandability to 64 bit addressing. TRON-based microprocessors include the 32-bit MN10400 by Matsushita, or a 32-bit Gmicro 300 from the combined efforts of Fujitsu, Hitachi, and Mitsubishi that I saw demonstrated. (The MN10400 has run floating point double precision calculations at 8.3MWIPS at 20MHz.) Oki Electric has developed a 0.8mu-m CMOS 32 bit TRON based micro (O32) containing 700,000 transistors, which will perform at 10MIPS at 33Mhz. The TRON Association also announced the development of CHIP64 (a 64-bit microprocessor). Hitachi has an IBM-PC bus board that allows its Integrated System Debugging Tool (ISDT) to run. ISDT is part of Hitachi's European based TRON project. Toshiba has a Intel-386 based operating system based on CTRON specifications. In the US, Interactive Systems Corp has ported Unix System V Release 3 to a Gmicro/200. This company has many years of experience in porting various versions of Unix to Intel, Motorola, and RISC processors and the development team found that the Gmicro/200 had some advantages that made hardware-dependent portions of the port relatively easy to implement (especially memory management and software generation system). Gmicro has a floating point unit (FPU). One interesting feature is that elementary functions are computed using the iterative "cordic" algorithm [C.W. Schelin, American Math Monthly, Vol 90, No. 5, May 1983, pp317-325]. Hewlett-Packard adapted a similar scheme for their pocket calculators. It has also been claimed that the Japanese industrial involvement is as much for fear of being left behind as from any direct interest in the project. Part of their reluctance stems from the historical Japanese approach to building custom software from scratch, rather than using standardized components. My own observation is that research activity is active although industrial commitments might be a bit tentative. As an example, here is a quote from the Mitsubishi Research Institute, "we...are conducting research into methods, based on the object-oriented approach, of defining application requirements, carrying out software design, and generating program code automatically. The object-oriented approach is especially geared to event-driven applications in real-time control fields so our immediate goal is to build prototype systems applying ITRON specifications." Matsushita has been most active in the adaptation of TRON into their product lines and has produced an educational system geared for the school market under the sponsorship of the Center for Educational Computing (CEC), an organization affiliated with MITI and the Ministry of Education and Culture. The potential adoption of TRON standards for school computers was one of the trade-related concerns. Some US researchers do not think much of TRON, saying it is nothing but a warmed-up version of the Motorola 68000 to make it a real time OS and extendable to 64 bit chip applications; in other words, not innovative technology. And some of the US vendors may feel that their real-time operating system kernels are superior to what could be done using ITRON. Sakamura feels that his viewpoint is rather different. TRON chips do not use RISC architecture. Sakamura believes that when floating point is required RISC speed drops off rapidly. Further he feels that to get the most performance and cost benefit it will be necessary to use specialized chips (ASIC), commenting that using a RISC chip for video is not as effective as using a special ASIC, and that it would cost far too much to use a RISC chip in a game computer or in what should be an inexpensive consumer product. He claims that a TRON-spec chip has functions that make it more suitable as an ASIC controller. Further, there is a family approach to TRON, something that RISC chips don't have (using the same architecture from 16 to 64 bits is not the RISC model). The TRON project is concerned about an architecture that is suited to systems with extremely large numbers of intelligent objects networked together, and compatibility is clearly necessary. A reasonable question is the relationship between TRON and other Japanese computer projects, such as the proposed New Information Processing Technology (see my report, NIPT.90, 26 Dec 1990). One major difference is that the funding for the latter is via the Japanese government; TRON is industrially financed. While TRON is looking toward futuristic uses of computers it is using silicon technology for chip design. Further, TRON has no provisions for new reasoning models, and concentrates more on the interface, communication, and collaboration issues. TRON is focused on uses of computers in very direct applications, while NIPT envisions much more complicated information processing requiring new models of what it means to think and compute. TRON is being propelled by scientists who imagine a city where many computers cooperate; NIPT is viewing a world where computers and people are synergistic. SUGGESTIONS. In the West many people are worried about TRON, because they fear that if Japanese electronic giants such as SONY, NEC, Fujitsu, Mitsubishi, and Matsushita someday adopt the TRON specification in their product standards, then these could eventually become the world-wide industry standard because of Japanese strength in the electronics and computer markets. Consequently, Western companies would then be forced to adopt TRON specifications in order to be competitive. There is another view however. As all TRON specs are published in English there is nothing to stop Western manufacturers from implementing a TRON operating system and marketing it to the Japanese. Computer Design (Feb 1, 1991) points out that TRON is only a specification, not real code. The fact that it can be implemented on many different levels appears to offer an opportunity for enterprising US software companies. They go on to remark that "if the US is so far ahead in software technology, it should be possible to create BTRON-based high-performance operating systems that would be strong competitors in the Japanese market". My own view is that we should concentrate on the scientific content of this project. There is no doubt that promotional literature about TRON is often vague and sometimes fails to differentiate between the future that will come anyway and the future that will come using TRON. Several of the Japanese academic computer science researchers I spoke to were also politely tentative about TRON. They state that CS research spans a broad range from highly theoretical to nuts-and-bolts extremely practical, and that TRON concentrates on quite practical applications. Nevertheless, there are many excellent ideas coming from the TRON community. Fueled by the Japanese ability in chip and other hardware design and manufacture members of this group have been aggressive and successful in building experimental systems. Western scientists are not going to be able to assess these unless they are more active participants in the project. Western standards organizations as well as research scientists should start paying serious attention to TRON specifications now, realizing that sooner than we anticipate, one of the Japanese giant electric firms may adopt TRON specifications. One way to begin would be to have serious US attendance at the 1991 TRON Symposium, which will be held on 26-27 November, 1991 in Tokyo. For more details contact Sakamura at the address given above. TRON BASED MICROPROCESSORS ------------------------------------------- TX1 Gmicro Gmicro Gmicro MN10400 O32 /100 /200 /300 Toshiba Mitsubishi Hitachi Fujitsu Matsushita Oki Number of Instructions 93 92 22(coproc) 102(basic) 93 103 22(coproc) 11(decimal) MMU X X O O X O Cache X 256 Inst. 1Kb Inst 2Kb Inst 1Kb Inst 1Kb Inst 128 Stack 2Kb Data 1Kb Data Number of Transistors 45 34 73 90 40 70 x 10,000 Process (CMOS) (in mu-m) 1.0 1.0 1.0 1.0 1.2 0.8 Packaging 155PGA 155PGA 135PGA 179PGA 144PGA 208PGA 152 OFP ------------------------------------------------------------------------- ITRON PRODUCTS----------------------------------------------------------- Specificantion OS CPU OS Vendor mu-ITRON MR7700 MELPS 7700 Series Mitsubishi Elec. MR3200 M32 Mitsubishi Elec. HI8-3X H8/300 Series Hitachi HI8 H8/500 Series (64Kb mem) Hitachi HI8-EX H8/500 Series (1Mb mem) Hitachi REALOS/7 F2MC-8 Series Fujitsu TR90 TLCS90 Series Toshiba ITRON1 RX116 V20/30 NEC HI68K 68000 Hitachi HI16 H16 Hitachi REALOS/286 80286 (protected mode) Fujitsu MR32 32032 Mitsubishi ITRON2 HI32 H32 Hitachi REALOS/F32 F32 Fujitsu IX101 TX1 Toshiba MR3210 M32 Mitsubishi ITRON/FILE HI68KA 68000 file mgmnt for HI68K Hitachi HI16A H16 file mgmnt for HI16 Hitachi HI32A H32 file mgmnt for HI32 Hitachi MR3200F M32 file mgmnt for MR3200 Mitsubishi Elec. MR3210 M32 file mgmnt for MR3210 Mitsubishi Elec. ------------------------------------------------------------------------- TRON ASSOCIATION MEMBERS AS OF MARCH 30, 1990. Aisin Seiki Co., Ltd. Alps Electric Co., Ltd. Amano Corporation AMD Japan Ltd. AMP (Japan), Ltd. Ampere Incorporated Ando Electric Co., Ltd. Anritsu Corporation Apple Computer Japan, Incl Asahi Kasei Microsystems Co., Ltd. Ascii Corporation AT&T Japan, Ltd. Aval Data Corp. Brother Industries, Ltd. Canon Inc. Casio Computer Co., Ltd. Central Information Center Co., Ltd. Chubu Electric Power Company, Inc. Computer Presence Corporation CSK Corporation Custom Technology Corporation Digital Electronics Corporation Digital Equipment Corporation Japan Electronics and Telecommunications Research Institute Elco International K.K. Fanuc Ltd. Ford Motor Company Fuji Electric Co., Ltd. Fuji Software Inc. Fuji Xerox Co., Ltd. Fuji Facom Corporation Fujitsu Limited Fujitsu Network Engineering Limited Garde Inc. Goldstar Software, Inc. Green Hills Software Inc. Hazama Corporation Hirose Electric Co., Ltd. Hitachi, Ltd. Hitachi Microcomputer Engineering, Ltd. Hitachi Software Engineering Co., Ltd. Hokkaido Information & Communication Co., Ltd. Hoshiden Electronics Co., Ltd. IBM Japan, Ltd. Ikegami Tsushinki Co., Ltd. Ines Corporation Intec Inc. Intel Japan k.k. Iwasaki Electronics Co., Ltd. Japan Air Lines Co., Ltd. Japan Aviation Electronics Industry, Ltd. Japan Direx Corporation Japan Radio Co., Ltd. Kawai Musical Instruments Manufacturing Co., Ltd. Keizo Limited Kohgaku-Sha Publishing Co., Ltd. Kohwa Joho Giken Inc. Kokusai Denshindenwa Co., Ltd. Kokusai Electric Co., Ltd. Kozu Systems Design Corporation KSD Corporation Kyocera Corporation Logic Systems International, Inc. Matsushita Communication Industrial Co., Ltd. Matsushita Electric Industrial Co., Ltd. Matsushita Electric Works, Ltd. Matsushita Electronics Corporation Meidensha Corporation Microboards, Inc. Micronics Co., Ltd. Microtec Research, Inc. Minolta Camera Co., Ltd. Misawa Homes Institute of Research and Development Co., Ltd. Mita Industrial Co., Ltd. Mitsubishi Electric Corporation Mitsubishi Electric Semiconductor Software Corporation Mitsubishi Research Institute, Inc. Mitsui Real Estate Development Co., Ltd. Morson Japan Motorola Inc. NEC Corporation Nihon Unisys, Ltd. Nippon Columbia Co., Ltd. Nippon-Data General Corporation Nippon Homes Corporation Nippon Koei Co., Ltd. Nippon System Kaihatsu Co., Ltd. Nippon Telegraph and Telephone Corporation Nippondenso Co., Ltd. Nissan Motor Co., Ltd. Nissin Electric Co., Ltd. NUK Corporation Northern Telecom Japan Inc. NTT Data Communications Systems Corporation NTT Software Corporation Oki Electric Cable Co., Ltd. Oi Electric Co., Ltd. Oki Electric Industry Co., Ltd. Olivetti Systems Technology Co. OMC, Incorporated Omron Corporation Omron Tateisi Software Co. Pasco Corporation Personal Media Corporation PFU Limited Plus Corporation Printing Machine Trading Co., Ltd. R&D Computer Co., Ltd. Ricoh Company, Ltd. Roland Corporation RSA Network Corporation Sanyo Electric Co., Ltd. Seiko Epson Corp. Seiko Instruments Inc. Seikosha Co., Ltd. Sharp Corporation Shimizu Corporation Shinko Electric Co., Ltd. Siemens AG Software Consultant Corp. Software Products and Systems Corporation Software Research Associates, Inc. Sony Corporation Sumitomo Electric Industries, Ltd. Sun Wave Industrial Co., Ltd. System Algo Co., Ltd. System V. Co. Texas Instruments Japan Limited Tokico Ltd. Tokyo Computer Service Co., Ltd. Tokyo Electric Power Co., Inc. Tosei Systems Co., Ltd. Toshiba Corporation Toto Ltd. Toyota Motor Corporation Uchida Yoko Co., Ltd. Uemura Giken Co., Ltd. Victor Company of Japan, Limited Wacom Co., Ltd. Wind River Systems, K.K. Yamaha Corporation Yamaichi Electric Mfg. Co., Ltd. Yamatake-Honeywell Co., Ltd. Yaskawa Electric Mfg. Co., Ltd. Yazaki Corporation Yokogawa Electric Corp. Yokogawa-Hewlett-Packard, Ltd. ------------------------------------------------------------------------- PROCEEDINGS OF THE SEVENTH TRON PROJECT SYMPOSIUM------------------------ TRON Project 1990 Open-Architecture Computer Systems Key Note Address Programmable Interface Design in HFDS K. Sakamura (Department of Information Science, Faculty of Science, University of Tokyo) Chapter 1: TRON Considerations of the Performance of a Real-Time OS A. Yokozawa, K. Fukuoka, K. Tamaru (Toshiba Corporation) Dynamic Stepwise Task Scheduling Algorithm for a Tightly-Coupled Multiprocessor ITRON N. Nishio, H. Takada, K. Sakamura (Department of Information Science, Faculty of Science, University of Tokyo) A Graphical Debugger for HI8 D. Wallace (Hitachi Europe Limited) HI8-3X: A mu-ITRON-Specification Realtime Operating System for H8/300 Series Microcontrollers M. Kobayakawa, T. Nagasawa, T. Shimizu, H. Takeyama (Microcomputer System Design Department, Semiconductor Development & Design Center, Hitachi, Ltd.) Chapter 2: BTRON Design Policy of the Operating System Based on the BTRON2 Specification K. Sakamura (Department of Information Science, Faculty of Science, University of Tokyo) A Study on a Hypermedia Editor on BTRON1 Specification Operating System K. Kajimoto, T. Nonomura (Kansai Information and Communiations Research Laboratories, Matsushita Electric Industrial Co., Ltd.) Chapter 3: CTRON CTRON Software Portability Evaluation T. Ohta, T. Terasaki (NTT Network Systems Development Center) T. Ohkubo, M. Hanazawa (NTT Electrical Communications Laboratories) M. Ohtaka (NEC Corporation) Portability Consideration of i386TM-Based Basic OS (OS/CT) K. Oda, Y. Izumi, H. Ohta, N. Shimizu, N. Yoshida (Toshiba Corporation) OS Subset Structure Achieving AP Portability H. Shibagaki, T. Wasano (NTT Network Systems Development Center) An Evaluation Method of Kernel Products Based on CTRON H. Kurosawa, O. Watanabe (Mitsubishi Electric Corporation) Y. Kobayashi (NTT Network Systems Development Center) Development of CTRON Operating System for Communication Processing M. Hatanaka, Y. Adachi, N. Shigeta, Y. Ohmachi, M. Ohminami (NTT Communications and Information Processing Laboratories) Chapter 4: CHIP (1) Implementation and Evaluation of Oki 32-bit Microprocessor 032 Y. Mori, Y. Haneda, Y. Arakawa, T. Mori, M. Kumazawa (OKI Electric Industry Co., Ltd.) Design Considerations of On-Chip-Type Floating-Point Units M. Suzuki, T. Kiyohara, M. Deguchi (Matsushita Electric Industrial Co., Ltd.) The Design Method of High Speed Cache Controller/Memory (CCM) for the GMICRO Family Microprocessors A. Yamada, H. Nakagawa, M. Hata, M. Satoh, K. Nishida (LSI Research and Development Laboratory, Mitsubishi Electric Corporation) T. Hiraki (Kita-Itami Works, Mitsubishi Electric Corporation) The Evaluation of M32/100's Bitmap Instructions Used in the Graphic Primitive M. Sakamoto, T. Shimizu, K. Saitoh (LSI Research and Development Laboratory, Mitsubishi Electric Corporation) Chapter 5: CHIP (2) Inline Procedures Boost Performance on TRON Architecture C. Franklin, C. Rosenberg (Green Hills Software, Inc.) A Forth Kernel for Gmicro1 H. Neugass (Microsystems Consultant) The Gmicro Microprocessor and the AT&T UNIX Operating System C. Reiher, W.P. Taylor (INTERACTIVE Systems Corporation) SRM32: Implementation of Symbolic ROM Monitor on GMICRO F32 Series Y. Kimura, H. Shida, S. Sasaki (Fujitsu Devices Inc.) H. Ito (Fujitsu Limited) Performance Evaluation of TOXBUS K. Okada (NTT Communication Switching Laboratories) M. Itoh, S. Fukuda (Fujitsu Limited) T. Hirosawa (Oki Electric Industry Co., Ltd.) T. Utsumi (Toshiba Corporation) K. Yoshioka (Hitachi Limited) Y. Tanigawa (Matsushita Electronics Corporation) K. Hirano (Mitsubishi Electric Corporation) Appendix: Additional Contributions Realtime OS TR90 Based on Micro-ITRON Specification K. Yamada, S. Takanashi (Toshiba Corporation) Y. Okada, M. Tamura (Toshiba Microelectronics Corporation) Communication Terminal for Heterogeneous Network Based on BTRON HMI N. Enoki, H. Oka, A. Yoneda, M. Ando (Kansai Information and Communications Research Laboratories, Matsushita Electric Industrial Co., Ltd.) Pitfalls on the Road to Portability J.D. Mooney (West Virginia University) Realization of the Micro-CTRON Kernel under pSOS+ A. Chao (Software Components Group, Inc.) ---------------------END OF REPORT---------------------------------------