Accelerators for Technical Computing: Is It Worth the Pain? A TCO Perspective

Abstract

Abstract. Nowadays, HPC systems emerge in a great variety including commodity processors with attached accelerators which promise to improve the performance per watt ratio. These heterogeneous architectures often get far more complex to employ. Therefore, a hardware purchase decision should not only take capital expenses and operational costs such as power consumption into account, but also manpower. In this work, we take a look at the total cost of ownership (TCO) that includes costs for administration and programming effort. From that, we compute the costs per program run which can be used as a comparison metric for a purchase decision. In a case study, we evaluate our approach on two realworld simulation applications on Intel Xeon architectures, NVIDIA GPUs and Intel Xeon Phis by using different programming models: OpenCL, OpenACC, OpenMP and Intel's Language Extensions for Offload.

Towards an accurate simulation of the crystallisation process in injection moulded plastic components by hybrid parallelisation

“…In these simulations, a reduction in memory consumption especially affects acquisition costs. However, an analysis considering total costs of ownership (as described by Wienke et al, 2013) will reveal a detailed insight.…”

Finite element numerical integration for first order approximations on multi-core architectures

“…• our example GPU, Tesla K20m, turned out to be the fastest processor in the comparison; the observations made by several authors [39,15,36] show that Xeon Phi processors applied to different scientific computing algorithms, although usually faster than standard Xeon CPUs, are usually slower than the recent HPC targeted GPUs…”