Fox Presentation July-Dec 1996 Second Collection of General Research Foils July -- December1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of July-Dec 1996 Research Foils This collects together Miscellaneous foils used in Research Presentations during second half of 1996 The first group of foils were used in trip to China July 12-28 1996 Status of "Classic" HPCC -- June1996 Overall Status: Success or Failure? http://www.npac.syr.edu/users/gcf/hpcc96status/index.html Presented during Trip to China July 12-28,1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Current Status 1996 We describe the structure of seven talks making up this review of HPCC from today to the Web and Petaflop performance in future Here we describe current status with HPCC in some sense both a failure and a great success This requires looking at hardware, software and the critical lack of commercial adoption of this technology We discuss COTS and trickle up and down technology strategies We describe education and interdisciplinary computational science in both simulation and information arenas Status of "Classic" HPCC -- June1996 Summary of MPP Hardware http://www.npac.syr.edu/users/gcf/hpcc96hardware/index.html Presented during Trip to China July 12-28,1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Hardware Status 1996 We describe basic technology driver -- the CMOS Juggernaut -- and some new approaches that could be important 10-20 years from now We describe from elementary point of view the basics of parallel(MPP) architectures We discuss current situation for tightly coupled systems -- convergence to distributed shared memory We discuss clusters of PC's/workstations -- MetaComputing Status of "Classic" HPCC -- June1996 Summary of MPP Software http://www.npac.syr.edu/users/gcf/hpcc96software/index.html Presented during Trip to China July 12-28,1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Software Status 1996 We start with an overall discussion of types of software environments and when they apply Data and Task Parallelism Coordination or Coarse Grain Software Integration Languages Data Parallel and Message Passing are still critical but the situation is confused by immaturity of parallel compilers We then discuss current work involving Xiaoming Li with HPF and the Parallel Runtime Compiler Consortium MetaComputing is an emerging field oof importance and we sketch our plans for MetaWeb Java threatens to change the ballgame! Status of "Classic" HPCC -- June1996 Summary of Grand/National Challenges, Applications, Acceptance by Industry http://www.npac.syr.edu/users/gcf/hpcc96appls/index.html Presented during Trip to China July 12-28,1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Applications Status 1996 We describe HPCC Applications starting with the many successes of Federal Grand Challenge Program in Government and Academic areas As a survey discovered, this does not translate into acceptance by industry We describe the trend to the the more broadly based National Challenges Industry has neither adopted the use of HPCC in their business operations nor has a viable software and systems industry (at high end) been created The resolution of "dilemma" of Industry v. National need in government and academia will underlie future programs Status of "Classic" HPCC -- June1996 Futures-1: Problem Solving Environments and Real Software in near future? http://www.npac.syr.edu/users/gcf/hpcc96pse/index.html Presented during Trip to China July 12-28,1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Futures 1:Problem Solving Environments Problem Solving Environments -- PSE's -- are seen in all fields from health care, education to engineering design of a new aircraft We illustrate with telemedicine Bridge concept And show in detail integration of NII and computation in ASOP -- next generation integrated manufacturing and design We give a couple of simple Web Computing Examples And outline NPAC's Web based strategy We describe needed enabling technologies and give a set of recommendations for progress coming from a panel led by John Rice of Purdue Status of "Classic" HPCC -- June1996 Futures-2: Petaflops and Real Software in 2007? http://www.npac.syr.edu/users/gcf/hpcc96petaflop/index.html Presented during Trip to China July 12-28,1996 Uses material from Peter Kogge -- Notre Dame Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Futures 2: PetaFlop in 2007! This describes some aspects of a national study of the future of HPCC which started with a meeting in February 1994 at Pasadena The SIA (Semiconductor Industry Association) projections are used to define feasible memory and CPU scenarios We describe hardware architecture with Superconducting and PIM (Processor in Memory possibilities) for CPU and optics for interconnect The Software situation is captured by notes from a working group at June 96 Bodega Bay meeting The role of new algorithms is expected to be very important Status of "Classic" HPCC -- June1996 Futures-3: Web Technology for HPCC? http://www.npac.syr.edu/users/gcf/hpcc96web/index.html Presented during Trip to China July 12-28,1996 Geoffrey Fox NPAC Syracuse University 111 College Place Syracuse NY 13244-4100 Abstract of HPCC Futures 3: Web Technology This describes Our Approach focussing on Integration of Information and Computing and concentrating on coarse grain functionality WebFlow : Dataflow (AVS) using Web with databases and numbercrunching MetaWeb : Metacomputing or rather cluster management using Web RSA Factoring was our first succesful example Financial Modelling will be an obviously important commercial application Java plays a critical role in high level user interfaces for visual programming, visualization of data and performance Web Interfaces to HPF will be particularly useful initially in education -- programming laboratories on the Web VRML is an interesting 3D datastructure Overall Remarks on the March to PetaFlops - I I find study interesting not only in its result but also in its methodology of several intense workshops combined with general discussions at national conferences Exotic technologies such as "DNA Computing" and Quantum Computing do not seem relevant on this timescale Note clock speeds will NOT improve much in the future but density of chips will continue to improve at roughly the current exponential rate over next 10-20 years Superconducting technology is currently seriously limited by no appropriate memory technology that matches factor of 100-1000 faster CPU processing Current project views software as perhaps the hardest problem Overall Remarks on the March to PetaFlops - II All proposed designs have VERY deep memory hierarchies which are a challenge to algorithms, compilers and even communication subsystems Major need for hig-end performance computers comes from government (both civilian and military) applications DoE ASCI (study of aging of nuclear weopens) and Weather/Climate prediction are two examples Government must develop systems using commercial suppliers but NOT relying on traditionasl industry applications to motivate So Currently Petaflop initiative is thought of as an applied development project whereas HPCC was mainly a research endeavour PSE Enabling Technologies I Collaborative computing technology Configuration control and human-in-the-loop (Computational steering) Computational geometry and grid generation Scalable algorithms Scalable solver libraries Parallel/Distributed computing --- metacomputing Fault tolerance and security Federated multi-media databases File system and I/O technologies Visualization including virtual reality, televirtuality etc. Interactive interface development (GUI) technologies Symbolic manipulation and automatic code generation Artificial intelligence and expert systems PSE Enabling Technologies II Performance monitoring and modeling "Low-level" virtual machine such as MPI, PVM etc. "Fine grain high-level" languages (C++, HPF etc.) Software engineering and coarse grain software (software bus) integration "Web-ware" and scripting middle-ware(Perl, Java, VRML, Python, etc.) Agent search and communication systems Wrapper technology for legacy systems and interoperability Interface specification support and information exchange protocols (such as CORBA, Opendoc, metadata and web standards) Object oriented software technology - object transport and management PSE templates and frameworks Possible Next Steps in HPCC PSE's Note that some steps -- such as Collaborative Environments will come from general NII activities Others such as integrated grid generators and geometry plus CFD solvers, and distributed scientific objects must come from HPCC The Federal High Performance Computing and Communication Initiative (HPCCI) Originally $2.9 billion over 5 years starting in 1992 and Rapidly growing Information technology component starting in 1994 and total budget became over $1 billion per year This drove race to Teraflop performance and is now OVER! The Grand Challenges Enabled by teraflop computers and important to economy or fundamental research Global warming - NOAA Oil reservoir and environmental simulation - DOE Structural and aerodynamic calculations - NASA Earth observing satellite - data analysis - NASA Human genome - NIH, DOE Quantum chromodynamics - Fundamental Physics Gravitational waves from black holes - Fundamental Physics Molecular modeling - Fundamental Chemistry What Happens now that HPCC Initiative is no longer in place? Most of the activities it started are ongoing! It achieved goal of Teraflop performance -- see Intel P6 machine at Sandia But it failed to develop a viable commercial base And although hardware peak performs at advertised rate, the software environment is poor This could be due to poor hardware as well as lack of sufficient resources to sustain software effort Academic Activities -- NSF Supercomputer centers -- are very healthy as much easier to put such codes on MPP as short in lifetime and lines of code Next initiatives -- based on PetaFlop goal -- will include a federal development as well as research component as can't assume "brilliant research" will be picked up by industry Components of This HPCC Presentation There are seven talks in this series: HPCC Status -- this talk -- Overall Technical and Political Status HPCC Today I -- MPP Hardware Architectures and Machines HPCC Today II -- Software HPCC Today III -- Applications -- Grand Challenges Industry HPCC Tomorrow I -- Problem Solving Environments HPCC Tomorrow II -- Petaflop (10^15 Operations per second) in the year 2007? HPCC Tomorrow III -- The use of Web Technology in HPCC Some HPCC Hardware Architectures and Their Status - I The future sees: 1:Mainly the relentless drive of CMOS juggernaut with Moore's law! 2:The growth of World Wide networked computers from settop boxes to MPP's 3:Processor in Memory (PIM) 4:Superconducting CPU's and interconnect (but no memory?) 5:Optical Interconnects 6:Quantum Computing (see Scientific American Oct 95 p140 by Seth Lloyd(MIT) 7:DNA or molecular computing We will discuss the first two here, the next three in "Petaflop futures" and I leave the last two to a future generation! Some HPCC Hardware Architectures and Their Status - II Today we see the following "CMOS Juggernaut" Architectures SIMD: No commercial or academic acceptance except for special purpose military (signal processing) and commercial(database indexing) applications Special Purpose: Such as GRAPE N-body machine which achieves a Teraflop today and a petaflop in a few years -- requires small memory and small CPU's MIMD Distributed Memory: Merging with Shared memory in tightly coupled systems Growing importance with World Wide Web and MetaComputing Shared Memory Tera Computer is isolated interesting attempt to build UMA system SGI (based on Stanford Work and merging Cray and SGI lines) and Convex will base high end on distributed shared memory which merges distributed and shared memory Physically distributed but logically NUMA shared memory Gordon Bell's SNAP Architecture - I Scalable Network (ATM) and Platforms (PC's running Windows 95) Gordon Bell's SNAP Architecture - II MetaComputing Built from PC's and ATM as commodity parts (COTS) Gordon Bell's SNAP Architecture - III The Computing World from Smart Card to Enterprise Server The Curious State of Parallel Software! MPI and PVM as Message Passing Systems are very healthy but the essential ideas are very old -- 10 to 20 years They are used because the systems work well (as relatively easy to build) and the users understand what to expect Perhaps hard to program this way but you know what you will get and it will do well if you do well! Parallel C++ is very confused with many standards HPF -- Data Parallel Fortran -- is a standard challenged by industry somewhat as they find compilers difficult and wish it was simpler Users find performance of HPF often disappointing and find it often is hard to predict what compiler will do HPF needs more "infrastructure investment". Not clear if it will make it!