The Transactions of the Korea Information Processing Society (한국정보처리학회논문지)

Korea Information Processing Society (한국정보처리학회)
권호 표지

An Improved Load Operand Referencing Scheme Using A Hybrid Predictor

혼합 예측기를 사용하는 효율적인 적재 명령어의 오퍼랜드 참조 기법

  • Published : 2000.07.01
Download PDF
⟨ Previous Next ⟩

Abstract

As processor's operational frequency increases and processors execute multiple instructions per cycle, the processor performance becomes more dependent on the load operand referencing latency and the data dependency. To reduce the operand fetch latency and to increase ILP by breaking the data dependency, we propose a value-address hybrid predictor using a reasonable size prediction buffer and analyse the performance improvement by the proposed predictor. Through the extensive simulation of 5 benchmark programs, the proposed hybrid prediction scheme accurately predicts 62.72% of all loads which are 12.64% higher than the value prediction scheme and show its cost-effectiveness compared to the address predition scheme. In addition, we analyse the performance improvement achieved by the stride management and the history of previous predictions.

Keywords

References

  1. B. Black, et al., 'Load Execution Latency Reduction,' Procs. of 1998 ICS, pp.29-35, 1998 https://doi.org/10.1145/277830.277842
  2. B. Calder, G. Reinman, D. Tullsen, 'Selective Value Prediction,' Procs. of 26th ISCA '99, pp.64-74, 1999 https://doi.org/10.1145/307338.300985
  3. Bob Cmelik and David Keppel, Shade, 'A Fast Instruction-Set Simulator for Execution Profiling,' Sigmetrics, ACM, pp.128-137, 1994 https://doi.org/10.1145/183018.183032
  4. C. Chi and C. Cheung, 'Hardware-Driven Prefetching For Pointer Data References,' Procs. of 1998 ICS, pp.377-384, 1998 https://doi.org/10.1145/277830.277924
  5. David L. Weaver and Tom Germond, The SPARC Architecture Manual, Prentice Hall, 1994
  6. Digital Equipment Corporation, Digital Semiconductor 21164 Alpha Microprocessor, Data Sheet, Digital Equipment Corporation, 1997
  7. Intel, Optimization's For Intel's 32- Bit Processors, available at http://ftp//download.intel.com, Intel Corporation, 1996
  8. Jose Gonzalez and Antonio Gonzalez, 'The Potential of Data Value Speculation to Boost ILP,' Proc. of 1998 ICS, pp.21-28, 1998 https://doi.org/10.1145/277830.277840
  9. L. Widigen, E. Sowadsky and K. McGrath, 'Eliminating Operand Read Latency,' Computer Architecture News, Vol.24, No.5, pp.18-22, 1996 https://doi.org/10.1145/242694.242708
  10. M. Bekerman, et al., 'Correlated Load-Address Predictor,' Proc. of 26th ISCA '99, pp.54-63, 1999 https://doi.org/10.1145/307338.300984
  11. M. Golden and Trevor Mudge, 'Hardware Support for Hiding Cache Latency,' CSE-TR-152-93, available at www.cs.umich.edu, U. of Michigan, 1993
  12. M. Lipasti and et al., 'Value Locality and Load Value Prediction,' Proc. of ASPLOS, pp.138-147, 1996 https://doi.org/10.1145/237090.237173
  13. Mikko. H. Lipasti, 'Value Locality and Speculative Execution,' PhD Dissertation, Carnegie Mellon University, 1997
  14. MIPS Technologies, MIPS R10000 Microprocessor User's Manual, Version 2.0, MIPS Technologies Inc., available at http//www.mips.com. 1996
  15. P. Ibanez, et al., 'Characterization and Improvement of Load/Store Cache-based Prefetching,' Proc. of the 1998. ICS, pp.369-376, 1998 https://doi.org/10.1145/277830.277921
  16. Tien-Fu Chen and Jean-Loup Baer, 'Effective Hardware-Based Data Prefetching for High-Performance Processors,' IEEE Transactions on Computer, Vol.44, No.5, pp.609-623, 1995 https://doi.org/10.1109/12.381947
  17. Tien-Fu Chen, 'Data Prefetching For High-Perfomance Processors,' PhD Dissertation, University of Washington, 1993
  18. 김홍준, 조경산, '오퍼랜드 참조 예측 캐쉬(ORPC)를 활용한 오퍼랜드 페치의 성능개선', 한국 정보처리학회 논문지, 제5권 제6호, pp.1652-1659, 1998