Measuring energy consumption for short code paths using RAPL

By Marcus Hähnel, Björn Döbel, M. Völp, H. Härtig. PERV, 2012

308 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Computer Vision and Pattern Recognition, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/23f4b96752fdb29b74b25f491e0df05a0701e6db

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

1. Wolfgang Maass
2. Robert A. Legenstein
3. Nicolas Papernot

In Computer Vision and Pattern Recognition

1. Wolfgang Maass
2. Robert A. Legenstein
3. Yiren Zhao
4. Nicolas Papernot
5. Robert D. Mullins
6. Guillaume Bellec
7. Ross Anderson
8. Ilia Shumailov

In Electrical and Electronic Engineering

1. Robert A. Legenstein
2. S. Furber
3. Wolfgang Maass
4. Michael Pfeiffer
5. Friedemann Zenke
6. J. Urgen Schmidhuber
7. Giacomo Indiveri
8. Herbert Jaeger

Citations in the corpus

• Memory-Efficient Deep Learning on a SpiNNaker 2 Prototype (2018, methodology): “We used the running average power limit (RAPL) energy sensor, which is a new feature provided by Intel since the Sandy Bridge micro-architecture (Hähnel et al., 2012), to measure the power consumption of the X86 CPU.”
• Sponge Examples: Energy-Latency Attacks on Neural Networks (2020, background): “RAPL has been thoroughly evaluated and found to reflect actual energy consumption, as long as the counters are not sampled too quickly [32, 33].”