C/C++ and Fortran

 View Only

POWER7 SPEC CPU2006 featuring automatic parallelization in XLC

By Archive User posted Thu May 20, 2010 08:48 PM

  

Originally posted by: rauls


With the recent publication of the SPEC CPU2006 scores of the POWER7-based p780 server, the IBM Power Systems have regained leadership on both the SPECint 2006 and SPECfp 2006 components of this industry benchmark suite.

In particular , the peak FP score of 71.5 is 20% higher than the previous best result.

One of the key features that have enabled this achievement is the use of automatic parallelization technology on several benchmarks. This not only highlights the advanced compilation and optimization technology in these compilers, but is also an indicator of the performance available on these systems when using parallel programming, for example through OpenMP.

Detailed results are available directly from spec.org:
http://www.spec.org/cpu2006/results/res2010q2/cpu2006-20100426-10753.html
http://www.spec.org/cpu2006/results/res2010q2/cpu2006-20100426-10752.html
SPEC® SPECint® and SPECfp® are registered trademarks of the Standard Performance Evaluation Corporation. Benchmark results stated above reflect results as of May 18, 2010. For the latest SPEC®CPU2006 benchmark results, visit www.spec.org.
1 comment
0 views

Permalink

Comments

Archive User
Fri May 28, 2010 09:55 AM

Originally posted by: M_F_Mokbel


Interesting results! Since it is not allowed to explicitly use OpenMP directives for parallelization according to SPEC rules. I wonder what does the auto-parallelizer in XLC uses to parallelize loops apart from auto-vectorization. Is it just through low level threading or does OpenMP directives get inserted implicitly. I'm actually interested in the scheduler used in cases where the workload characterization is not available.