Alea

Abstract

Energy efficiency is becoming increasingly important, yet few developers understand how source code changes affect the energy and power consumption of their programs. To enable them to achieve energy savings, we must associate energy consumption with software structures, especially at the fine-grained level of functions and loops. Most research in the field relies on direct power/energy measurements taken from on-board sensors or performance counters. However, this coarse granularity does not directly provide the needed fine-grained measurements. This article presents ALEA, a novel fine-grained energy profiling tool based on probabilistic analysis for fine-grained energy accounting. ALEA overcomes the limitations of coarse-grained power-sensing instruments to associate energy information effectively with source code at a fine-grained level. We demonstrate and validate that ALEA can perform accurate energy profiling at various granularity levels on two different architectures: Intel Sandy Bridge and ARM big.LITTLE. ALEA achieves a worst-case error of only 2% for coarse-grained code structures and 6% for fine-grained ones, with less than 1% runtime overhead. Our use cases demonstrate that ALEA supports energy optimizations, with energy savings of up to 2.87 times for a latency-critical option pricing workload under a given power budget.

Characterization of HPC workloads on an ARMv8 based server under relaxed DRAM refresh and thermal stress

“…For example, Figure 4a shows how the number of CEs induced by kmeans(8) changes in time for DRAM operating under 2.283 S T RE F P at 60 • C. We can indentify two different phases in this benchmark: at the first phase the average number of CEs per second is less than 20, while at the second phase this number achieves almost 300 per second. By profiling and analysis of this benchmarks [16,17], we found that the first phase corresponds to the I/O phase of kmeans where the input data for this benchmark is retrieved from a file. While at the second phase, the benchmark processes the input data reading and writing to DRAM intensively which explains the high error rate obtained at this phase.…”