An evaluation of the many-core Longtium SP computer system

The utilization wall, which lead to the Dark Silicon phenomenon, indicates that the energy efficiency will be the primary constraint in micro-processor architecture design in the near future. Since the stream processing natively has the high energy efficiency feature, we proposed the many-core Longtium stream processor (SP) architecture, which includes a coarsegrained reconfigurable array (CGRA), a simplified star-mesh static NoC and software managed stream memory system. In this paper, we demonstrate the evaluation methodology for Longtium SP. The methodology includes the FPGA prototyping system, baseline processor, software tool-chains, golden model and benchmarks. The FPGA prototyping system show that Longtium SP could averagely speedup the program by 11.6X and improve the energy efficiency by 177X, after taking the chip area into account it still could speedup the program by 5.8X compared to the dual-issue baseline RISC processor.

[1]  Karthikeyan Sankaralingam,et al.  Dark Silicon and the End of Multicore Scaling , 2012, IEEE Micro.

[2]  Xiaoya Fan,et al.  Design and performance analysis of one 32-bit dual issue RISC processor for embedded application , 2008, 2008 9th International Conference on Solid-State and Integrated-Circuit Technology.

[3]  Peter Mattson,et al.  A programming system for the imagine media processor , 2002 .

[4]  Jeffrey R. Diamond,et al.  An evaluation of the TRIPS computer system , 2009, ASPLOS.

[5]  Michael Bedford Taylor,et al.  Is dark silicon useful? Harnessing the four horsemen of the coming dark silicon apocalypse , 2012, DAC Design Automation Conference 2012.

[6]  Scott Rixner,et al.  Stream Processor Architecture , 2001 .

[7]  Henry Hoffmann,et al.  Evaluation of the Raw microprocessor: an exposed-wire-delay architecture for ILP and streams , 2004, Proceedings. 31st Annual International Symposium on Computer Architecture, 2004..

[8]  William J. Dally,et al.  Architectural Support for the Stream Execution Model on General-Purpose Processors , 2007, 16th International Conference on Parallel Architecture and Compilation Techniques (PACT 2007).