ppt课件-coarse grain reconfigurable architectures（粗粒可重构体系结构）.pptVIP

Software or Configware? About the Digital Divide of Parallel Computing Reiner Hartenstein TU Kaiserslautern Preface The ?havenots“ Software to Configware Migration HPC HPC Embedded Computing The wrong Roadmap Configware Engineering Dual Machine Paradigms Speed-up Examples Final Remarks data are moved around by software path of least resistance*:avoiding a paradigm shift … understand only this parallelism solution: Embedded Computing HPC Embedded Computing The wrong Roadmap Configware Engineering Dual Machine Paradigms Speed-up Examples Final Remarks History of Machine Models History of Machine Models the hardware / Software Chasm: Growth Rate of Embedded Software typical CS graduates: the ?havenots“ the current CS mind set is based on the Submarine Model Hardware / Configware / Software Partitioning skills urgently needed By the way ... Dominance of the Submarine Model ... CS Education the wrong Roadmap HPC Embedded Computing the wrong Roadmap Configware Engineering Dual Machine Paradigms Speed-up Examples Final Remarks Completely wrong roadmap Completely wrong mind set 3rd machine model became mainstream Configware Engineering Supercomputing (HPC) Embedded Computing The wrong Roadmap Configware Engineering Dual Machine Paradigms Speed-up Examples Final Remarks de facto Duality of RAM-based platforms .... the brain hurts From Software to Configware Industry benefit from RAM-based 2nd paradigm Nick Tredennick’s Paradigm Shifts explain the differences Compilation: Software vs. Configware  Flowware: not new data streams*: not new 1980: data streams (Kung, Leiserson: systolic arrays) 1989: data-stream-based Xputer architecture 1990: rDPU (Rabaey) 1994: Flowware Language MoPL (Becker et al.) 1995: super systolic array (rDPA) + DPSS tool (Kress) 1996+: Stream-C language, SCCC (Los Alamos), SCORE, ASPRC, Bee (UC Berkeley), ... 1996+: streaming languages (Stanford et al.) 1996: configware / software partitioning compiler (Be