Evaluation of the Energy Performance of Dense Linear Algebra Kernels on Multi-core and Many-Core Processors

We evaluate the power consumption of three key kernels from several tuned dense linear algebra libraries on three general-purpose multi-core processors and a graphics processor, representative of the state-of-the-art in computing technology. The results of this study provide basic insights on the energy scalability of multi- and many-core designs and multi-threaded software as the building blocks of future EXAFLOPS systems.

[1]  FengWu-chun,et al.  The Green500 List , 2007 .

[2]  R. Clint Whaley,et al.  Achieving accurate and context‐sensitive timing for code optimization , 2008, Softw. Pract. Exp..

[3]  Enrique S. Quintana-Ortí,et al.  Exploiting the capabilities of modern GPUs for dense matrix computations , 2009 .

[4]  Gene H. Golub,et al.  Matrix computations (3rd ed.) , 1996 .

[5]  James Demmel,et al.  LU, QR and Cholesky Factorizations using Vector Capabilities of GPUs , 2008 .

[6]  Thomas Ludwig,et al.  Simulation of power consumption of energy efficient cluster hardware , 2010, Computer Science - Research and Development.

[7]  Feng Pan,et al.  Analyzing the Energy-Time Trade-Off in High-Performance Computing Applications , 2007, IEEE Transactions on Parallel and Distributed Systems.

[8]  Thomas Ludwig.  Editorial for the First International Conference on Energy-Aware High Performance Computing , 2010, Computer Science - Research and Development.

[9]  M. Horowitz,et al.  Low-power digital design , 1994, Proceedings of 1994 IEEE Symposium on Low Power Electronics.

[10]  Robert H. Halstead,et al.  Matrix Computations , 2011, Encyclopedia of Parallel Computing.

[11]  Thomas Ludwig,et al.  Editorial for the second international conference on energy-aware high performance computing , 2011, Computer Science - Research and Development.

[12]  Susanne Albers,et al.  Energy-efficient algorithms , 2010, Commun. ACM.

[13]  Constantine Bekas,et al.  A new energy aware performance metric , 2010, Computer Science - Research and Development.