论文信息 - Developing a Compiler for the XeonPhi

Developing a Compiler for the XeonPhi

The XeonPhi is a highly parallel x86 architecture chip made by Intel. It has a number of novel features which make it a particularly challenging target for the compiler writer. This paper describes the techniques used to port the Glasgow Vector Pascal Compiler (VPC) to this architecture and assess its performance by comparisons of the XeonPhi with 3 other machines running the same algorithms.

[1] J. Paul Siebert,et al. Parallel Stereo Vision Algorithm , 2012, MARC@RWTH.

[2] Jan Paul Siebert,et al. C3D™: a Novel Vision-Based 3-D Data Acquisition System , 1995 .

[3] W. Paul Cockshott,et al. Vector pascal reference manual , 2002, SIGP.

[4] Dr. Martyn J. Corden,et al. Consistency of Floating-Point Results using the Intel ® Compiler or Why doesn ’ t my application always give the same answer ? , 2010 .

[5] Thomas Raymond Turner,et al. Vector pascal: a computer programming language for the fps-164 array processor , 1987 .

[6] George Chrysos,et al. Intel® Xeon Phi coprocessor (codename Knights Corner) , 2012, 2012 IEEE Hot Chips 24 Symposium (HCS).

[7] Peter Cooper. Porting the Vector Pascal Compiler to the Playstation 2 , 2005 .

[8] Emre Kultursay,et al. Compiler-Based Data Prefetching and Streaming Non-temporal Store Generation for the Intel(R) Xeon Phi(TM) Coprocessor , 2013, 2013 IEEE International Symposium on Parallel & Distributed Processing, Workshops and Phd Forum.

[9] Youssef Gdura,et al. A new parallelisation technique for heterogeneous CPUs , 2012 .

[10] W. Paul Cockshott,et al. Orthogonal parallel processing in vector Pascal , 2006, Comput. Lang. Syst. Struct..