Apparatus, method, and system for improving power, performance efficiency by coupling a first core type with a second core type

作者: Edson Borin , Youfeng Wu , Shiliang Hu , Mauricio Breternitz , Cheng C. Wang

DOI:

关键词:

摘要: An apparatus and method is described herein for coupling a processor core of first type with co-designed second type. Execution program code on the monitored hot sections are identified. Those optimize execution core, such that upon subsequently encountering those sections, optimized executed co- designed core. When executing code, may be in low-power state to save power or other parallel. Furthermore, multiple threads cold pipelined while pipeline achieve maximum performance.

参考文章(24)
Thomas M. Conte, Andrew Wolfe, Mapping of computer threads onto heterogeneous resources ,(2010)
Gad S Sheaffer, Ronny Ronen, Murali M. Annavaram, Doron Orenstein, Hong Wang, Edward Grochowski, John Shen, Method and apparatus for varying energy per instruction according to the amount of available parallelism ,(2004)
Keith Farkas, Parthasarathy Ranganathan, Norman P. Jouppi, Robert N. Mayo, Dynamically selecting processor cores for overall power efficiency ,(2003)
Premanand Sakarda, Jay Heeb, Mark N. Fullerton, Hong-Yi Chen, Sehat Sutardja, Dynamic core switching ,(2008)
Brian V. Belmont, Animesh Mishra, James P. Kardach, Dynamic core swapping ,(2014)
Timothy Sherwood, Brad Calder, Gregory Hamerly, Erez Perelman, Method and apparatus for identifying similar regions of a program's execution ,(2003)
Douglas Carmean, Herbert Hum, Rajesh Kumar, Eric Sprangle, Distribution of tasks among asymmetric processing elements ,(2014)
G.M. Silberman, K. Ebcioglu, An architectural framework for supporting heterogeneous instruction-set architectures IEEE Computer. ,vol. 26, pp. 39- 56 ,(1993) , 10.1109/2.214441