The Salishan Conference

The Salishan Conference on High-Speed Computing was founded in 1981 as a means of getting experts in computer architecture, languages, and algorithms together to

  • improve communications,
  • develop collaborations,
  • solve problems of mutual interest, and
  • provide effective leadership in the field of high-speed computing.

Organized by a Tri-Lab committee and hosted by the Association for High-Speed Computing, the conference convenes annually at the Salishan Lodge in Gleneden Beach, Oregon.

Attendance at the conference is by invitation. Attendees are from national laboratories, academia, government, and private industry. To preserve the level of interaction and discussion, conference attendance is kept to about 150.

Conference Material by Year


Theme: Energy-Centric HPC: Systems and Applications

Keynote Address: Potential Roles for High Speed Computing in the Obama Administration, Vic Reis, U.S. Department of Energy

Session 1: Energy Applications I

Session 2: Energy Applications II

Working Dinner/Speaker

Session 3: Energy Efficient Computing I

Session 4: Energy Efficient Computing II

Session 5: HPC Platforms Update


Theme: HPC in the Era of Ubiquitous Parallelism: Multicore and Hybrid Architectures

Keynote Address: Multicore Meets Exascale: The Catalyst for a Software Revolution, Kathy Yelick, Lawrence Berkeley National Laboratory

Session 1: Processor Architecture Roadmap

  • Exascale-the Next Great Challenge, Peter Kogge, University of Notre Dame, and William Harrod, DARPA
  • Exascale Ambitions What, me worry? : ΔS > kB ln2, William J. Camp, Intel Corporation
  • The Role of Accelerated Computing in the Multi-core Era, Charles Moore, AMD*
  • Why CPU's Have to Evolve: From Homogeneous to Heterogeneous Chips, a Brief Overview, Michael Paolini, IBM Corporation

Session 2: System Software

  • Systems Software Challenges and Strategies for the Petascale/Exascale Era, Fred Johnson, DOE Office of Advanced Scientific Computing Research
  • The Role of Compilers and Programming Languages for Client-Side Multicore Systems, Vikram Adve, University of Illinois, Urbana-Champaign
  • Quad-core Catamount and R&D in Multi-core Lightweight Kernels, Kevin Pedretti, Sandia National Laboratories, SAND number 2008-1725A
  • Petascale Communication is not Business as Usual, Al Geist, Oak Ridge National Laboratory

Working Dinner/Speaker

  • Multicore: Hey, Wait a Minute?, Dan Reed, Microsoft

Session 3: Applications

  • SciDAC and the Path Toward Exascale, Walter Polansky, Office of Advanced Scientific Computing Research, Office of Science
  • Kinetic Plasma Modeling with VPIC: Status and Future Plans on Hybrid Architectures, Brian Albright, Los Alamos National Laboratory, LA-UR-08-2557
  • Coping with Petascale Architectures, Bronis R. de Supinski, Lawrence Livermore National Laboratory, LLNL-PRES-403033
  • Auto-tuned Optimization of Scientific Kernels on Leading Multicore Systems, Leonid Oliker, Lawrence Berkeley National Laboratory

Session 4: Programming Models/Environment

  • Programming Models and Languages for High Performance Computing, Marc Snir, University of Illinois, Urbana-Champaign
  • Toward an Open and Unified Model for Heterogeneous and Accelerated Multicore Computing, Catherine Crawford, IBM Corporation, LA-UR-08-3500
  • Transactional Memory for a Modern Microprocessor, Marc Tremblay, Sun Microsystems, Inc.

Session 5: System Architectures

  • The Institute for Advanced Architectures and Algorithms, Sudip Dosanjh, Sandia National Laboratories, and Jeff Nichols, Oak Ridge National Laboratory, SAND 2008-2204P
  • Sequoia Architectural Requirements, Matt Leininger, Lawrence Livermore National Laboratory, LLNL-PRES-403472
  • The Cray Roadmap to Cascade, John Levesque, Cray, Inc.
  • Moore, More Cores, and More Application Performance, Darren Kerbyson, Los Alamos National Laboratory, LA-UR-08-2847

Random Access

  • Hybrid MPI + OpenMP Issues, Rolf Rabenseifner, University of Stuttgart
  • Is Dynamic Memory Management Dynamic Enough? Jeff Keasler, Lawrence Livermore National Laboratory, LLNL-PRES-403029
  • A 192K Process BG/L Simulation of Laser-plasma Interactions, Steve Langer, Bert Still, Denise Hinkel, Bruce Langdon, and Ed Williams, Larwrence Livermore National Laboratory, LLNL-PRES-403016.
  • Redesigning Dense Linear Algebra Software for ManyCore and Beyond, Jack Dongarra, University of Tennessee and Oak Ridge Natinal Laboratory
  • Asian HPC Update (Japan, China, India), David Kahaner, Asian Technology Information Program (ATIP)
  • Nuclear Energy Advanced Modeling and Simulation (NEAMS), Alex Larzelere, Advanced Simulation and Modeling Office of Nuclear Energy, Fuel Cycle Management (NE-5)
  • Whole System Performance Instrumentation, Patrick Worley, Oak Ridge National Laboratory


Theme: Confidence in HPC Predictive Simulations

Keynote Address: Verification and Validation for Turbulence Mixing , James Glimm, Stony Brook University

Session 1: Overview of ASC V&V Program

Session 2: V&V and UQ Analysis Methodologies

Working Dinner/Speaker

  • Beyond Games: Challenges and Opportunities for Leveraging High-Volume Technologies in HPC, H. Peter Hofstee, IBM Systems and Technology Group

Session 3: Predictive Capability Case Studies

Session 4: V&V of the Computational Infrastructure

Session 5: Advanced Architecture Systems


Theme: Data Intensive Computing

Keynote Address: HPC I/O and file Systems: Is Everyone Out to Get Us? (pdf), Gary Grider, Los Alamos National Laboratory

Session 1: Data Intensive Applications

Session 2: Analysis Methodologies

Working Dinner/Speaker

Session 3: I/O Issues, Technology and Directions

Session 4: File Systems

Session 5: Archival Storage Systems


Theme: Paths to Petaflops

Opening Address: Modeling Tsunamis with ASC Codes, Galen Gisler, LANL

Keynote Address: Petaflop-Scale Computing (pdf), Dimitri Kusnezov, DOE, NNSA

Session 1: Exploring Requirements for Petaflops Computing

Session 2: Architecture Models for Petaflops Systems

Session 3: Other Designs for Petaflops Architectures

Session 4: System Software Tools for Petaflops Systems

Session 5: Productivity Issues for Petaflops Systems


Theme: Applications, Algorithms, Architectures – Past Accomplishments, Current Issues, and Future Trends

Keynote Address: The Next Quarter-Century at Salishan (pdf), Burton J. Smith, Chief Scientist, Cray, Inc.

Session 1: 25 Years of Problems, Challenges, and Successes

Session 2: Applications

Session 3: Algorithms

Session 4: Architectures

Session 5: The Next 10 Years in High-Performance Computing


Theme: Extreme Scale Computing

Keynote Address: Moore’s Law – A Time for Reflection and Refraction; Petaflops in 2009/10, Steven J. Wallach, Chiaro Networks, Ltd.

Session 1: Critical Components in Extreme Scale System Architectures

Session 2: System Architectures of Large HPC Platforms

Session 3: Applicability of Linux in HPC Systems

Session 4: User Perspectives of Current HPC Systems

Session 5: Homeland Defense Applications