1 / 4

Session 1: GPU: Graphics Processing Units

Session 1: GPU: Graphics Processing Units. Miriam Leeser / Northeastern University HPEC Conference 15 September 2010. Manycore is Here to Stay. Current accelerators of choice: Multicore processors (2, 4 or 8 processors on a chip) Graphics processing units (GPUs) NVIDIA, AMD

mayes
Download Presentation

Session 1: GPU: Graphics Processing Units

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Session 1:GPU: Graphics Processing Units Miriam Leeser / Northeastern University HPEC Conference 15 September 2010

  2. Manycore is Here to Stay • Current accelerators of choice: • Multicore processors (2, 4 or 8 processors on a chip) • Graphics processing units (GPUs) • NVIDIA, AMD • Cloud computing • The future: • Manycore processors (16, 32, 64 processors on a chip) • More flexible GPUs • AMD Fusion, NVIDIA Fermi, Intel Knight’s Corner • Cloud computing on a chip • Intel’s Single Chip Cloud Computer

  3. This Morning’s Talks • Invited speakers: • Accelerating Mechanical Computer Aided Engineering (MCAE) applications with GPUs • Dr. Robert Lucas, USC ISI • Thinking outside the Tera-Scale box • PiotrLuszczek, University of Tennessee at Knoxville • GPUs: • Sparse Matrix Algorithms on GPUs and their Integration into SCIRun • Devon Yablonski, Northeastern University • Benchmark Evaluation of Radar Processing Algorithms on Graphics Processor Units (GPUs) • Scott Sawyer, Lockheed Martin • Failing In Place for Low-Serviceability Infrastructure Using High-Parity GPU-Based RAID • Anthony Skjellum , University of Alabama at Birmingham

  4. The Future • Manycore computing and GPUs: • From dozens to thousands of cores on a single chip • SIMD: Single Instruction Multiple Data • Vector processing • Network on a chip • A mix of styles • CPU, GPU, … • Programming techniques, languages, methodology • ?

More Related