An array control unit for high performance SIMD arrays

Although arrays of SIMD PEs can be built with very high operating frequencies, problems exist in keeping the array busy. The inherent mismatch between host and array makes it difficult to maintain high array utilization: either the rate of instruction issue is very low or PE data locality is compromised, having the same effect. Our solution is based on an array control unit (ACU) design that expands macro instructions in two stages, first by data tile and then into microinstructions. The expansion itself solves the issue problem; decoupling the expansion modalities maintains data locality. Several issues involving host/ACU interaction need to be resolved to effect this solution.

[1]  Y. Fujita,et al.  A 10 GIPS SIMD processor for PC-based real time vision applications -architecture, algorithm implementation and language support , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[2]  Tom Blank,et al.  The MasPar MP-1 architecture , 1990, Digest of Papers Compcon Spring '90. Thirty-Fifth IEEE Computer Society International Conference on Intellectual Leverage.

[3]  Benjamin Bishop,et al.  Three dimensional graphics algorithms on the Micro-Grain Array Processor. II , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[4]  A. Broggi,et al.  A dedicated image processor exploiting both spatial and instruction-level parallelism , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[5]  T. J. Fountain.  The design of highly-parallel image processing systems using nanoelectronic devices , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[6]  David E. Schimmel,et al.  Issues in the Design of High Performance SIMD Architectures , 1996, IEEE Trans. Parallel Distributed Syst..

[7]  Martin C. Herbordt,et al.  Processor/memory/array size tradeoffs in the design of SIMD arrays for a spatially mapped workload , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[8]  P. P. Jonker,et al.  The CC/IPP, an MIMD-SIMD architecture for image processing and pattern recognition , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[9]  Charles Sodini,et al.  System design for pixel-parallel image processing , 1996, IEEE Trans. Very Large Scale Integr. Syst..

[10]  Martin C. Herbordt,et al.  Making a dataparallel language portable for massively parallel array computers , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[11]  Shin'ichiro Okazaki,et al.  A 64 parallel integrated memory array processor and a 30 GIPS real-time vision system , 1995, Proceedings of Conference on Computer Architectures for Machine Perception.

[12]  Masatoshi Ishikawa,et al.  Vision chip architecture using general-purpose processing elements for 1 ms vision system , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[13]  Changhee Lee,et al.  A general purpose SliM-II image processor , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[14]  Changhee Lee,et al.  A General Purpose SKM-I1 Image Processor , 1997 .

[15]  Martin C. Herbordt,et al.  A system for evaluating performance and cost of SIMD array designs , 1999, Proceedings. Frontiers '99. Seventh Symposium on the Frontiers of Massively Parallel Computation.

[16]  Martin C. Herbordt,et al.  Experimental Analysis of Some SIMD Array Memory Hierarchies , 1995, ICPP.

[17]  Mary Jane Irwin,et al.  Computer vision on the MGAP , 1993, 1993 Computer Architectures for Machine Perception.

[18]  Michael Bolotski,et al.  Abacus - a reconfigurable bit-parallel architecture for early vision , 1996 .