No abstract

In high-performance computing (HPC), significant effort is invested in research and development of novel memory technologies. One of them is Spin Transfer Torque Magnetic Random Access Memory (STT-MRAM)-byteaddressable, high-endurance non-volatile memory with slightly higher access time than DRAM. In this study, we conduct a preliminary assessment of HPC system performance impact with STT-MRAM main memory with recent industry estimations. Reliable timing parameters of STT-MRAM devices are unavailable, so we also perform a sensitivity analysis that correlates overall system slowdown trend with respect to average device latency. Our results demonstrate that the overall system performance of large HPC clusters is not particularly sensitive to main-memory latency. Therefore, STT-MRAM, as well as any other emerging non-volatile memories with comparable density and access time, can be a viable option for future HPC memory system design. CCS Concepts •Computer systems organization → Processors and memory architectures; •Hardware → Non-volatile memory; •Computing methodologies → Massively parallel and high-performance simulations;

Energy consumption is by far the most important contributor to HPC cluster operational costs, and it accounts for a significant share of the total cost of ownership. Advanced energy-saving techniques in HPC components have received significant research and development effort, but a simple measure that can dramatically reduce energy consumption is often overlooked. We show that, in capacity computing, where many small to medium-sized jobs have to be solved at the lowest cost, a practical energy-saving approach is to scale-in the application on large-memory nodes. We evaluate scaling-in; i.e. decreasing the number of application processes and compute nodes (servers) to solve a fixedsized problem, using a set of HPC applications running in a production system. Using standard-memory nodes, we obtain average energy savings of 36%, already a huge figure. We show that the main source of these energy savings is a decrease in the node-hours (node hours = #nodes × exe time), which is a consequence of the more efficient use of hardware resources.Scaling-in is limited by the per-node memory capacity. We therefore consider using large-memory nodes to enable a greater degree of scaling-in. We show that the additional energy savings, of up to 52%, mean that in many cases the investment in upgrading the hardware would be recovered in a typical system lifetime of less than five years.

No abstract

No abstract