A performance analysis of multicore computer architectures Michel Schelske
2 Parallel Algorithms for Multicore Benchmarking10. Apr Structure 1.Observations & Theory 2.Problems 3.Solution
3 Parallel Algorithms for Multicore Benchmarking10. Apr Observation I clock rate cores performance 3-4 GHz cores frequency performance Multi performance
4 Parallel Algorithms for Multicore Benchmarking10. Apr Observation II programthread partitioning granularity
5 Parallel Algorithms for Multicore Benchmarking10. Apr Theory Optimum depends on hardware and the problem to be solved granularity performance coarse-grainfine-grain
6 Parallel Algorithms for Multicore Benchmarking10. Apr Example
7 Parallel Algorithms for Multicore Benchmarking10. Apr Observation III Optimum depends on hardware and the problem to be solved coarse-grain granularity performance fine-grain
8 Parallel Algorithms for Multicore Benchmarking10. Apr Observation IV Optimum depends on hardware and the problem to be solved granularity performance coarse-grainfine-grain
9 Parallel Algorithms for Multicore Benchmarking10. Apr The problems Granularity is only one performance parameter. Find the optimal parallelization parameters with respect to – the algorithm – the computer architecture
10 Parallel Algorithms for Multicore Benchmarking10. Apr Our Solution hardware core operating system Application Profiler Benchmark
11 Parallel Algorithms for Multicore Benchmarking10. Apr Thank you for your attention
12 Parallel Algorithms for Multicore Benchmarking10. Apr Result calculation of prime numbers on a computer with two Intel Xeon Singlecore CPUs