Running a number of industry level test and experiments, CFMS undertook an independent product evaluation of IBM Power8 System S824L, which provided factual outcomes that demonstrate product capability in the growing technical computing and HPC arena.
CFMS undertook an independent product technology evaluation of IBM Power8 System S824L, assessed for readiness and utilisation in technical computing and HPC workload.
Product evaluation covered a number of industry standard tests including system performance for technical computing and engineering workloads, quality of the toolchain (compilers, environment, etc.) and integrated energy efficiency and scale. The testing programme also reviewed total cost of ownership (TCO) and management, business benefits and industry relevance.
Providing insight from an actual user perspective, services consist of benchmarking a product against manufacturer performance through to product validation where CFMS input into product design and development, to confirm the proposed approach. CFMS team of IT systems and industry specialists in modelling and simulation set-up testing environments and run cases for vendors and industry end users.
SMT8 can be reconfigured while the system is running, allowing for the SMT configuration to be optimised for each simulation run, or even at each simulation job step. POWER8 provides greater Non-Uniform Memory Access (NUMA) control, allowing the optimisation of the process layout on the hardware topology.
Observed is an ongoing trend within the HPC landscape for heterogeneous clusters, rather than attempting a ‘best-fit’ homogenous configuration. These heterogeneous clusters may combine standard CPU only compute nodes with either ‘high-memory’ nodes, or with compute nodes equipped with GPGPU or other accelerators. Tools like xCAT (eXtreme Cloud Administration Toolkit) and IBM Spectrum Scale (formerly IBM GPFS) can be used to deploy and run mixed x86 and Power systems (equipped with NVIDIA GPGPUs) to accelerate specific workloads, while providing a consistent experience for end users. Porting existing CUDA software to run on the S824L’s K40 GPUs was trivial, primarily due to the common interfaces provided by the CUDA toolkits on x86 and POWER8.
System performance assessment focused on compute rather that I/O, as typical HPC workloads would exercise processors with storage being implemented as a shared parallel file system. The performance was measured by running zCFD on a standard aerospace test case from NASA. The latest CUDA 7.0rc Toolkit from NVIDIA was utilised, together with the gcc 4.9 providing OpenMP thread based parallelism and OpenMPI 6.5 providing MPMD parallelism.
To view the full results of the IBM Power8 System S842L Benchmarking and Testing, click here.