25th November 1993 /parallel/documents/inmos/archive-server The complete contents of the INMOS (US) email archive server (8 Megabytes of files) provided by Hugh Thomas of INMOS US. Many thanks to Hugh for this. 24th November 1993 /parallel/software/emacs-lisp/occam-mode.el Improved version of occam2 mode for GNU emacs by Jared Saia with "pretty good automatic indentation of occam code by syntax, auto capitilization of Occam keywords, the ability to comment regions of occam code in a way that the compiler accepts and some other random things." /parallel/software/emacs-lisp/old-occam-mode.el Older version of above. 23rd November 1993 [Note to net.police: this is for informational purposes, not advertising] /parallel/documents/vendors/APR/ProductInfo/ Product information from ftp site netcom.com:/pub/forge/ProductInfo /parallel/documents/vendors/APR/ProductInfo/README Contents /parallel/documents/vendors/APR/ProductInfo/dpf.man.txt FORGE Magic Distributed Memory Parallelizer dpf manual page /parallel/documents/vendors/APR/ProductInfo/dpf_datasheet.txt FORGE Magic Distributed Memory Parallelizer dpf Datasheet /parallel/documents/vendors/APR/ProductInfo/forge90_datasheet.txt Interactive FORGE 90 Parallelizer Datasheet /parallel/documents/vendors/APR/ProductInfo/forgex_datasheet.txt FORGE Explorer Interactive Fortran Browser forgex Datasheet /parallel/documents/vendors/APR/ProductInfo/magic_datasheet.txt FORGE Magic Automatic Parallelizers Intro Datasheet /parallel/documents/vendors/APR/ProductInfo/news.txt Latest News from APR (November 1993) /parallel/documents/vendors/APR/ProductInfo/pricing.txt (nearly empty) /parallel/documents/vendors/APR/ProductInfo/xhpf.man.txt FORGE Magic HPF Parallelizer xhpf manual page /parallel/documents/vendors/APR/ProductInfo/xhpf_datasheet.txt FORGE Magic HPF Parallelizer xhpf Datasheet /parallel/conferences/concurrency-in-computational-logic Call for papers for Workshop on Concurrency in Computational Logic being held on 13th Decemeber 1993 at the Department of Computer Science, City University, London, United Kingdom. Deadline for papers: 30th November 1993. /parallel/documents/vendors/formal-systems/fdr.announce Announcement of availaility of academic licences for FDR (Failures Divergence Refinement) tool. The distribution is available to educational institutions for a nominal media charge. It is a refinement checking tool for CSP. /parallel/conferences/hpcn-europe-94-update Updated call for papers and details of the European Conference on High-Performance Computing and Networking (HPCN Europe) 1994 being held from April 18-20 1994 at the Sheraton Hotel & Towers Munich, Germany. Deadlines: Papers: 30th November 1993; Posters: 30th November 1993. /parallel/conferences/alp-plilp94.ps.Z /parallel/conferences/alp-plilp94.tex /parallel/conferences/alp-plilp94.txt Call for papers for ALP'94 and PLILP'94 - the Fourth International Conference on Algebraic and Logic Programming and the Sixth International Symposium on Programming Language Implementation and Logic Programming being held from 14th-16th September 1994 at Madrid, Spain. Deadlines: Papers: 28th February 1994; Acceptance: 10th May 1994; Final copy: 20th June 1994. /parallel/user-groups/hpff/hpff-II Call for attendance at the High Performance Fortran Forum II Kickoff Meeting being held from 13th-14th January 1994 at the Wyndham Greenspoint Hotel, Houston, Texas, USA. /parallel/documents/misc/X3H5-standard-comments Call for comments on the proposed ANSI standard: X3.252.199x, Parallel Processing Model for High Level Programming Languages /parallel/conferences/euromicro-workshop-on-par+dist-processing Revised details of registration and programme of the 2nd Euromicro Workshop on Parallel and Distributed processing from January 26-28th 1994 at the University of Malaga, Spain. /parallel/conferences/ieee-workshop-fault-tolerant-par-dist-systems Call for papers for 1994 IEEE Workshop on Fault-Tolerant Parallel and Distributed Systems being held from 13th-14th June 1994 at College Station, Texas, USA sponsored by the IEEE Computer Society Technical Committee on Fault-Tolerant Computing in cooperation with IFIP Working Group 10.4 and Texas A&M University and in conjunction with FTCS-24. Deadlines: Papers: 18th February 1994; Acceptance: 15th April 1994; Revised Paper: 15th August 1994; Panel Proposals: 15th March 1994 22nd November 1993 /parallel/reports/gatech/ossurvey.announcement Announcement of report below and call for comments by Bodhisattwa Mukherjee (bodhi@cc.gatech.edu> /parallel/reports/gatech/ossurvey.ps.Z A Survey of Multiprocessor Operating System Kernels (DRAFT) by Bodhisattwa Mukherjee (bodhi@cc.gatech.edu>, Karsten Schwan and Prabha Gopinath ABSTRACT: Multiprocessors have been accepted as vehicles for improved computing speeds, cost/performance, and enhanced reliability or availability. However, the added performance requirements of user programs and functional capabilities of parallel hardware introduce new challenges to operating system design and implementation. This paper reviews research and commercial developments in multiprocessor operating system kernels from the late 1970's to the early 1990's. The paper first discusses some common operating system structuring techniques and examines the advantages and disadvantages of using each technique. It then identifies some of the major design goals and key issues in multiprocessor operating systems. Issues and solution approaches are illustrated by review of a variety of research or commercial multiprocessor operating system kernels. /parallel/conferences/par-high-perf-apps Call for attendance at the Conference and Tutorial on Parallel High-Performance Applications being held from 15th-17th December 1993 at KTH, Stockholm, Sweden. Registration before 1st December 1993. /parallel/reports/announcements/update-propogation-in-galactica-net Announcement of Technical Report available for FTP: "Update Propagation in the Galactica Net Distributed Shared Memory Architecture" by A. Wilson, R. LaRowe, R. Ionta, R. Valentino, B. Hu, P. Breton and P. Lau of Center For High Performance Computing of Worcester Polytechnic Institute, Marlborough, MA, USA. /parallel/documents/vendors/APR/magic-dm-auto-parallelization-tools Announcement of MAGIC series of automatic parallelizing pre-compilers, FORGE Magic/DM for distributed memory systems and clustered workstations, and FORGE Magic/SM for shared memory parallel systems. /parallel/documents/vendors/APR/parallelizing-products Details of APR parallelizing products (FORGE 90, xHPF77) /parallel/documents/misc/DIMACS-implementation-challenge-III Call for participation in the Third DIMACS Internations Algorithm Implementation Challenge. This will take place between November 1993 and September 1994 by carrying out research projects related to the problem areas specified to present research papers at a DIMACS workshop in October 1994. /parallel/documents/misc/faster-messaging-shmem-multiprocs-video Details of Video tape on "Faster Messaging in Shared-Memory Multiprocessors", Optimizing Memory-Based Messaging for Scalable Shared Memory Multiprocessor Architectures, 8th September 1993, 71 minutes by Bob Kutter and David Cheriton. /parallel/user-groups/cray/conference-spring-1994 Call for papers and details of Cray Users Group Spring '94 conference being held from 14th-18th March 1994 at San Diego, California, USA. Deadlines: Papers and Posters: 10th December 1994. 19th November 1993 /parallel/documents/vendors/intel/cmu-intel-iWarp-ftp-site Details of FTP site for publications relating to the Carnegie Mellon/ Intel SSD built iWarp parallel computers. /parallel/faqs/cm5-intel-mailing-lists Details of NAS/NASA mailing lists for the Thinking Machines CM5, Intel Paragon and Intel iPSC/860 machines. [ Added new area: /parallel/documents/benchmarks and moved other ] [ benchmark-related documents here. ] /parallel/documents/benchmarks/PAR93-SMP Announcement of PAR93 -- a benchmark suite designed to measure cache based RISC SMP system performance using well behaved codes that parallelize automatically. /parallel/software/announcements/distributed-c-development-environment Details of the Distributed C Development Environment developed at Technische Universitaet Muenchen, Germany by a group under Prof. Dr. J. Eickel. Available for networks of UNIX computers. Runs on Sun SPARCstations (SunOS), Hewlett Packard workstations (HP/UX), IBM workstations (AIX), Convex supercomputers (ConvexOS), IBM Workstations (AIX) and homogeneous and heterogeneous networks of the systems as mentioned above. Public domain software. /parallel/courses/parasoft-par-prog-course-linda "Introduction to Parallel Programming with Linda" course given by ParaSoft Corporation on 9th December 1993 at Florida State University, Tallahassee, Florida, USA immediately after the Cluster Workshop '93. 18th November 1993 /parallel/conferences/app-par-geoscience Call for papers for the European Geophysical Society XIX General Assembly symposium session EGS2 on Applications of Parallel Processing in Geoscience being held from 25th-29th April 1994 in Gernoble, France. Deadlines: Abstracts: 1st January 1994; Young Scientist and East European Awards: 15th December 1994. /parallel/conferences/spaa94 Call for papers for the 6th Annual ACM Symposium on Parallel Algorithms and Architectures being held from 27th-29th June 1994 at Cape May, New Jersey, USA sponsored by the ACM Special Interest Groups for Automata and Computability Theory (SIGACT) and Computer Architecture (SIGARCH) and organized in cooperation with the European Association for Theoretical Computer Science (EATCS). Deadlines: Papers: 21st January 1994; Acceptance: 15th March 1994; Final-copy: 8th April 1994. /parallel/books/introductory-parallel-books A summary of responses to a query about introductory books for parallel programming/algorithms. Replies by David Bader and Jason Moore 12th November 1993 /parallel/books/ios/nrat/ "Networks, Routers & Transputers" book as compressed PostScript edited by M.D. May, P.W. Thompson and P.H. Welch. /parallel/books/ios/nrat/Overview Overview of the book - READ THIS FIRST and pick a chapter /parallel/books/ios/nrat/Introduction.ps.Z Introduction /parallel/books/ios/nrat/Chapter1.ps.Z Transputers and Routers: Components for Concurrent Machines /parallel/books/ios/nrat/Chapter2.ps.Z The T9000 Communications Architecture /parallel/books/ios/nrat/Chapter3.ps.Z DS-Links and C104 Routers /parallel/books/ios/nrat/Chapter4.ps.Z Connecting DS-Links /parallel/books/ios/nrat/Chapter5.ps.Z Using Links for System Control /parallel/books/ios/nrat/Chapter6a.ps.Z /parallel/books/ios/nrat/Chapter6b.ps.Z Models of DS-Link Performance Parts 1 and 2 /parallel/books/ios/nrat/Chapter7.ps.Z Performance of C104 Networks /parallel/books/ios/nrat/Chapter8.ps.Z General Purpose Parallel Computers /parallel/books/ios/nrat/Chapter9.ps.Z The Impl. of Large Par. Database Machines on T9000 and C104 Nets /parallel/books/ios/nrat/Chapter10a.ps.Z /parallel/books/ios/nrat/Chapter10b.ps.Z /parallel/books/ios/nrat/Chapter10c.ps.Z /parallel/books/ios/nrat/Chapter10d.ps.Z A Generic Architecture for ATM Systems Parts 1 to 4 /parallel/books/ios/nrat/Chapter11.ps.Z An Enabling Infr. for a Distributed Multimedia Industry /parallel/books/ios/nrat/Appendices.ps.Z Appendices A to D 9th November 1993 [ Added Working Group documents and papers and reports for the ] [ IEEE Draft Std P1355 - Standard for Heterogeneous InterConnect (HIC) ] [ (INMOS's DS Links) as well as datasheets for IMS C101 and IMS C104 ] [ These are the documents on the INMOS UK ftp site, cleaned up ] /parallel/documents/inmos/ieee-hic/copper.ps.Z Long Distance Differential Transmission of DS Links over Copper Cable by Stefan Haas, Xinjian Liu and Brian Martin of CERN. 34 pages. 981195 bytes uncompressed. /parallel/documents/inmos/ieee-hic/data/C101.ps.Z IMS C101 parallel DS-Link adaptor - Preliminary Datasheet 52 pages. 1679579 bytes uncompressed. /parallel/documents/inmos/ieee-hic/data/C104.ps.Z IMS C104 packet routing switch - Preliminary Datasheet 58 pages. 5834421 bytes uncompressed. /parallel/documents/inmos/ieee-hic/draftd0.0.ps.Z IEEE Draft Std P1355 - Standard for Heterogeneous InterConnect (HIC) (Low Cost Low Latency Scalable Serial Interconnect for Parallel System Construction) of 7th October 1993. Version D0.0 80+ pages. 2306156 bytes uncompressed. /parallel/documents/inmos/ieee-hic/fiber.ps.Z GP-MIMD T9000 Fiber Optic Link Extensions - Report on 850nm Fiber Optic Transceiver by Stefan Haas of ECP Division, CERN. 22 pages. 417516 bytes uncompressed. /parallel/documents/inmos/ieee-hic/pressrel.txt Press release about IEEE P1355 Working Group. July 1993. /parallel/documents/inmos/ieee-hic/roster.txt Roster of Working Group members /parallel/documents/inmos/ieee-hic/wg19oct.txt Working group minutes of 19th October 1993 /parallel/documents/inmos/ieee-hic/wg1sep.txt Working group minutes of 1st September 1993 /parallel/documents/inmos/ieee-hic/wg22jun.txt Working group minutes of 22nd June 1993 8th November 1993 /parallel/journals/jcse-par-alg-arch Call for papers for the Journal of Computer and Software Engineering Special Issue on Parallel Algorithms and Architectures to be published around January 1995. Deadline: 1st May 1994. /parallel/conferences/ieee-workshop-par-dist-simulations Call for papers for ACM/IEEE/SCS 8th Workshop on Parallel and Distributed Simulation being held from 6th-8th July 1994 at the University of Edinburgh, Scotland, UK. Deadlines: Papers: 1st December 1993; Notification: 1st March 1994; Camera-ready copy: 15th April 1994. /parallel/faqs/linux-and-transputers Summary of current state of Linux and transputers by Michael Haardt <(michael)u31b3hs@pool.informatik.rwth-aachen.de>, author of the assembler and server package. /parallel/courses/applications-of-par-programming Applications for High Performance Computers 14th-17th December 1993 given by Telmat Informatique in Soultz, France. Deadline for registration: 6th December 1993. /parallel/conferences/raps-workshop-par-benchmarks Call for attendance for the RAPS Open Workshop on Parallel Benchmarks and Programming Models being held from 7th-8th December 1994 at Chilworth Manor Conference Centre, Southampton, UK. /parallel/conferences/massive-parallelism Call for papers for 2nd International Workshop on Massive Parallelism: Hardware, Software and Applications being held from 3rd-7th October 1994 and organized by: Istituto di Cibernetica, Naples, Italy. Deadlines: Manuscripts: 1st February 1994; Notification: 31st April 1994; Camera-ready papers: 31st May 1994. /parallel/books/mit/concurrent-oo.announce Announcement of book: "Research Directions in Concurrent OOP" edited by Gul Agha, Peter Wegner and Aki Yonezawa. /parallel/conferences/podc94 Call for papers for 1994 ACM Symposium on Principles of Distributed Computing (PODC) being held from 14th-17th August 1994 at Los Angeles, California, USA. Deadlines: Abstracts: 4th February 1994; Acceptance: 15th April 1994; Camera-ready copy: 10th May 1994. /parallel/software/linux/assembler/server.tz /parallel/software/linux/assembler/asld.tz Updated versions. [ Added release notes for Mentat 2.6. Note: No binary distribution ] [ (executables) is included here due to the licensing restrictions. It ] [ must be obtained from the source site directly and not distributed from ] [ your site. See README for details. ] /parallel/software/virginia/mentat/README.2.6 New features in version 2.6 /parallel/software/virginia/mentat/Release26_notes.ps.Z Release notes for version 2.6 2nd November 1993 /parallel/conferences/ipps94-parallel-io-workshop Call for papers for the 2nd Annual Workshop on I/O In Parallel Computer Systems being held at IPPS (International Parallel Processing Symposium) 94. The workshop will be held on 26th April 1994 at Hotel Regina, Cancun, Mexico. Deadlines: Paper: 31st January 1994; Notification: 15th March 1994; Camera-ready copy: 7th April 1994. /parallel/documents/vendors/cray/cray-cs64000.announcement Announcement of the CRAY CS64000 by Cray Research, Inc. /parallel/software/announcements/pablo Details of PABLO: a system for the collection, display, and analysis of parallel program performance data developed by Prof. Daniel A. Reed's research group at the University of Illinois at Urbana-Champaign, USA. /parallel/faqs/classification-of-parallel-algorithms Summary of responses to a query about classifying parallel algorithms by Marion Wittmann /parallel/conferences/hicss28-par-and-dist-computing Call for Minitrack Propsals for the Software Technology Track focussing on Parallel and Distributed Computing: Theory, Systems and Applications at HICSS-28 (28th Hawaii International Conference on System Sciences) being held at Maui, Hawaii, USA from 3rd-6th January 1995. [A minitrack is either a half day or a full day of technical sessions. All sessions are conducted in a workshop-like setting and participants often participate in several different tracks] Deadlines: Proposals: 26th November 1993; Notification: 31st December 1993. /parallel/conferences/int-conf-par-processing-1994 Call for papers for the 1994 International Conference on Parallel Processing (23rd Annual Conference) being held from 15th-19th August 1994 at The Pennsylvania State University, Illinois, USA. Deadlines: PAPERS: 10th January 1994; Acceptance: 20th March 1994; TUTORIALS: 1st March 1994. [ Updated WoTUG, NATUG and WoTUG/Japan user group details: ] /parallel/user-groups/transputer-user-groups Contact addresses for known world transputer user groups /parallel/user-groups/natug/committee.doc North American Transputer User Group (NATUG) committee /parallel/user-groups/wotug/committee.doc WoTUG committee members details /parallel/user-groups/wotug/constitution.doc The new constitution of WoTUG, approved at the Sheffield meeting and first AGM. /parallel/user-groups/wotug/sig-chairs.doc WoTUG Special Interest Group (SIG) chairs /parallel/user-groups/wotug/minutes/1993-03-31 Minutes of the first World occam and Transputer User Group (WoTUG) Annual General Meeting (AGM) held during the 16th WoTUG Technical Meeting, on 31 March 1993 at Earnshaw Hall, University of Sheffield, UK. Minutes Secretary: Julie Clarke (University of Sheffield) /parallel/user-groups/wotug-japan/committee.doc World occam and Transputer User Group Japan committee /parallel/papers/twente Papers from researchesas at the Mechatronics Research Centre Twente and Control Laboratory, Department of Electrical Engineering, University of Twente, Netherlands. /parallel/papers/twente/fft.ps.Z "A Generalized FFT algorithm on transputers" by Herman Roebbers, University of Twente, The Netherlands; Peter Welch, University of Kent at Canterbury, UK; Klaas Wijbrans, Van Rietschoten & Houwens, The Netherlands. ABSTRACT: "A Generalized alogrithm has been derived for the execution of the Cooley-Tukey FFT algorithm on a distributed memory machine. This algorithm is based on an approach that combines a large number of butterfly operations into one large process per processor. The performance can be predicted from theory. The actual algorithm has been implemented on a transputer array, and the performance of the implementation has been measured for various sizes of the complex input vector. It is shown that the algorithm scales linearly with the number of transputers and the problem size." /parallel/papers/twente/linxback.ps.Z "The Twente LINX backplane" by M.H. Schwirtz, K.C.J. Wijbrans, A.W.P. Bakkers, E.P. Hoogzaad and R.Bruis of the Mechatronics Research Centre Twente and Control Laboratory, Department of Electrical Engineering, University of Twente, Netherlands. ABSTRACT: "The design of a control system is not finished with the derivation of the necessary control algorithms. When the controller is implemented in a digital computer, the system designed has to schedule all control and calculation tasks within the sampling interval of the system. Higher sampling frequencies often improve system performance. On the other hand, more sophisticated control algorthims require more computing time thus reducing the obtainable sampling frequencies. Therefore, it is important to minimise the overhead of sampling and communications. This paper describes a transputer-baseed I/O system fulfilling this requirement and shows how the sampling with this system is done." /parallel/papers/twente/pga-kernel.ps.Z "Post-Game Analysis on Transputers - Development of a Measurement Kernel" by J.P.E Sunter, E.C. Koenders and A.W.P. Bakkers, Mechatronics Research Centre Twente and Control Laboratory, Department of Electrical Engineering, University of Twente, Enschede, Netherlands. ABSTRACT: "Int this paper Post-game analysis, a method for alloation of processes on an arbitrary network of processors, is investigated. Contrary to other methods, this method is not based on a-prior information. To generate allocations it uses heuristics and measurements obtained during program execution using a previous allocation. This has the advantage that it is not influenced by inaccuracies in a-priori information. This methods needs a few iterations to come to a good allocation. If the number of iterations is small, this method is a good repalcement for the computationally intensive deterministic methods". /parallel/papers/twente/pga-performance.ps.Z "Performance of Post-Game Analysis on Transputers" by J.P.E. Sunter and A.W.P. Bakkers, Mechatronics Research Centre Twente and Control Laboratory, Department of Electrical Engineering, University of Twente, Enschede, Netherlands. ABSTRACT: "In this paper the performance of a Post-Game Analysis system is studied. For this purpose several simple cases have been designed. These consist of a number of processes which have to be distributed over a network of transputers. The final distributions for these cases are compared to the optimal nes, which can be easily derived for these simple cases. Performance measures consider include the number of iterations required to reach the final distribution, and the overhead caused by monitoring the program behaviour." /parallel/papers/twente/prio-sched.ps.Z "Cooperative Priority Scheduling in Occam" by J.P.E. Sunter, K.C.J. Wijbrans and A.W.P. Bakkers, Mechatronics Research Centre Twente, Electrical Engineering Department, University of Twente, Enschede, Netherlands. ABSTRACT: "In this paper a scheduler for variable priority scheduling is presented. This scheduler assumes that the processes being scheduled cooperate with the scheduler. This cooperation introduces some latencey in the scheduling of the processes. Analytic experssions describing the effect of this latency are dervied. A variable priority scheduler was implemented and results from actual program executions are given. This results show that the scheduler can be used to schedule algorithms with simple sequential processes with sample frequencies not higher than 2 kHz." /parallel/papers/twente/virtual.ps.Z "Virtual Channel Generator - VCG" by Johan P.E. Sunter and A.W.P. Bakkers, Mechatronics Research Centre Twente, University of Twente, Enschede, Netherlands. ABSTRACT: "This paper deals with a novel way of implementing communication layers for transputer networks. In the past, the limitation of the number of links of a transputer to four has led to the development of many network layers and operating system kernels that provide topology independent routing. Programming multitransputer systems is facilitated this way, by providing a transparent communications service through kernel calls or special communication channels. Because these layers are designed as operating systems or library functions, the complete layer is always added to the transputer system. Thus the layers introduce considerable overhead for applications using irregularly sized data transfer, a high communication/computation ratio or large packet sizes. Especially for real-time applications this is not acceptable. This paper describes a different approach that combines design-time flexibility with run-time efficiency. Instead of always adding the same general purpose kernel for the provision of a transparent communications service, a dedicated kernel is generated for each transputer. This is possible, because a-priori knowledge can be extracted at compile time from the application processes that are loaded onto each transputer. This knowledged consists of the communication size on the channels and of the the toplogy of the application process. The network generator, the Virtual Channel Generator (VCG), is sufficiently smart to recognize situations in which it is not necessary to add network processes. As a result, the network layer is optimially adapted to the requirements of the specfic application." /parallel/papers/twente/while-sched.ps.Z "List Scheduling While Loops on Transputers" by Johan P.E. Sunter and A.W.P. Bakkers, Mechatronics Research Centre Twente, University of Twente, Enschede, Netherlands. ABSTRACT: "List scheduling is a well known tool for scheduling sequential programs on parallel machines. However, these sequential programs are not allowed to contain loops. In this paper this restriction is removed. A list scheduler is presented which allows the sequential programs to contain nested while loops." /parallel/conferences/wotug17 Call for papers for the 17th World occam and Transputer User Group (WoTUG) Technical Meeting being held from 11th-13th April 1994 at the University of Bristol, UK. Deadlines: Extended abstracts: 11th November 1993; Notification: Mid December 1993; Camera ready copy: 10th January 1994.