The Energy Measurement Library provides a simple API to extract energy consumption data from hardware power counters. It is designed to simplify energy consumption experimentation by presenting a unified interface for different kinds of hardware.
The following devices are currently supported:
-
Dummy devices for testing purposes
-
Intel CPUs starting from Sandy Bridge (through the Running Average Power Limit interface)
-
Recent Nvidia Tesla and Quadro GPUs (through NVIDIA Management Library)
-
Intel Xeon Phi MICs (through the Intel Manycore Platform Software Stack (3.x+), from the host device)
-
Schleifenbauer PDUs (through the Schleifenbauer socket API)
-
Odroid-XU3 integrated sensors (through linux sysfs)
-
Poznań Supercomputing and Networking Center Labee XML Interface (through its REST API)
EML automatically discovers necessary libraries and available devices at runtime.
Doxygen-generated documentation can be found at:
EML is released under the GPLv2 license.
Required:
- libconfuse, for configuration parsing
Optional:
- NVML, for Nvidia GPU support
- MPSS 3.x+, for Intel MIC support
- libcrypto, for Schleifenbauer PDU support (RC4 encryption)
- libxml2, for Labee support
- libcurl, for Labee support
Universidad de La Laguna, High Performance Computing group [email protected]
Homepage: http://cap.pcg.ull.es
This work was supported by the Spanish Ministry of Education and Science through TIN2011-24598 and TIN2016-78919-R project, through the FPU program and the Spanish network CAPAP-H. It also has been partially supported by EU under the COST programme Action IC1305, ‘Network for Sustainable Ultrascale Computing (NESUS)’ and the European COST Action CHIPSET. Additional support was received through project ProID2017010130 and the grant TESIS2017010134, cofinanced by the Ministry of Economy, Industry, Commerce and Knowledge of Canary Islands and the European Social Funds (ESF).