GPU Architectures &
Computing
VU Ezio Bartocci
Vienna University of Technology
Overview • ECTS: 6.0 • Web: http://ti.tuwien.ac.at/cps/teaching/courses/gpu_architectures_and_computing/
• Lecturer: Ezio Bartocci • Lectures: Tuesday/Thursday 10-12 am • Institute: E182 Institute of Computer Engineering (Library - 3rd Floor) • Language: English • Enrollment Deadline: 7 March • Prerequisites:
• VU Algorithmen und Datenstrukturen 1 • VU Algorithmen und Datenstrukturen 2 • 182.695 LU Digital Design and Computer Architecture • 182.709 UE Operating Systems • 351.015 VU Signals and Systems 1 • 389.055 VU Signals and Systems 2
Aim of the course The objectives of this course are:
• gaining understanding of GPU computer architecture
• getting familiar with GPU programming environments
• Implementing programs solving problmes that would classically have been run on supercomputers
The requirements for this course are:
• Basic notions of Computer Architecture
• Good knowledge of C/C++ programing is expected
Main Topics
The main topics of the course are:
• GPU Architectures
• CUDA Programming Model
• OpenCL Programming Model
• GPU Computing: Case Studies
• Optimizing GPU performances
Our equipment (1)
Our Rack is now equipped with:
• 4 NVIDIA Kepler K10
• 8 GPU with 12288
• 16 Teraflops in sigle precision
• 1 Teraflops in double precision
• 64 GB of RAM available on CPU • 32 GB of RAM available on GPU
Our equipment (2)
NVIDIA Carma (2):
• Tegra 3 quad-core ARM A9 processor,
• Quadro 1000M GPU with 96 cores (good for 270 single-precision GFlops)
• one Gigabit Ethernet interface
EMBEDDED SOLUTION
Some GPU Applications
3D Model of a Pig Heart (Fenton-‐Karma 3V Model)
3D Model of a Mouse Heart (Fenton-‐Karma 3V Model)
Minimal Model in Four State Variables (4V)
Suggested Books CUDA by Example: An IntroducPon to General-‐Purpose GPU Programming Jason Sanders, Edward Kandrot, Addison Wesley
Programming Massively Parallel Processors: A Hands-‐on Approach David B. Kirk, Wen-‐mei W. Hwu, Morgan Kaufmann
CUDA ApplicaPon Design and Development Rob Farber, Morgan Kaufmann
GPU CompuPng GEMS Emerald EdiPon Wen-‐Mei w. Hwu, Morgan Kaufmann
Heterogeneous CompuPng with OpenCL Benedict Gaster, Lee Howes, David R. Kaeli, Perhaad Mistry, Dana Schaa, Morgan Kaufmann
Suggested Books
Grading
• Student Presentation 30%
• Lab Protocol 20% • Code 50%
Top Related