(Back to Session Schedule)

The 14th Asia and South Pacific Design Automation Conference

Session 2D  Special Session: EDA Acceleration Using New Architectures
Time: 13:30 - 15:35 Tuesday, January 20, 2009
Location: Room 416+417
Organizer: Damir A. Jamsek (IBM Corp., United States)

2D-1 (Time: 13:35 - 14:15)
Title(Invited Paper) Programming Multicore SIMD Architectures
AuthorReiji Suda (The University of Tokyo, Japan)
AbstractA General introduction to SIMD and multi-threading programming.

2D-2 (Time: 14:15 - 14:55)
Title(Invited Paper) Designing and Optimizing Compute Kernels on Nvidia GPU's
Author*Damir Jamsek (IBM Research Division, United States)
KeywordGPU, NVIDIA
AbstractThe availability of high performance compute capability in NVIDIA GPUs has expanded their use in CAD environments. We will describe the basic compute models including host/device programming models, device multi-thread programming models, as well optimization and performance tuning techniques

2D-3 (Time: 14:55 - 15:35)
Title(Invited Paper) Parallelizing Fundamental Algorithms such as Sorting on Multi-core Processors for EDA Acceleration
AuthorMasato Edahiro (System IP Core Research Laboratories, NEC Corporation/Department of Computer Science, University of Tokyo, Japan)