BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset
The BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset adds node-level energy consumption data from watt-meters to the primary sweep of the BUTTER - Empirical Deep Learning Dataset. This dataset contains energy consumption and performance data from 63,527 individual experimental runs spanning 30,582 distinct configurations: 13 datasets, 20 sizes (number of trainable parameters), 8 network "shapes", and 14 depths on both CPU and GPU hardware collected using node-level watt-meters. This dataset reveals the complex relationship between dataset size, network structure, and energy use, and highlights the impact of cache effects.
BUTTER-E is intended to be joined with the BUTTER dataset (see "BUTTER - Empirical Deep Learning Dataset on OEDI" resource below) which characterizes the performance of 483k distinct fully connected neural networks but does not include energy measurements.
Complete Metadata
| @type | dcat:Dataset |
|---|---|
| accessLevel | public |
| bureauCode |
[
"019:20"
]
|
| contactPoint |
{
"fn": "Charles Tripp",
"@type": "vcard:Contact",
"hasEmail": "mailto:charles.tripp@nrel.gov"
}
|
| dataQuality |
true
|
| description | The BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset adds node-level energy consumption data from watt-meters to the primary sweep of the BUTTER - Empirical Deep Learning Dataset. This dataset contains energy consumption and performance data from 63,527 individual experimental runs spanning 30,582 distinct configurations: 13 datasets, 20 sizes (number of trainable parameters), 8 network "shapes", and 14 depths on both CPU and GPU hardware collected using node-level watt-meters. This dataset reveals the complex relationship between dataset size, network structure, and energy use, and highlights the impact of cache effects. BUTTER-E is intended to be joined with the BUTTER dataset (see "BUTTER - Empirical Deep Learning Dataset on OEDI" resource below) which characterizes the performance of 483k distinct fully connected neural networks but does not include energy measurements. |
| distribution |
[
{
"@type": "dcat:Distribution",
"title": "BUTTER-E Paper",
"format": "08151v1#S3",
"accessURL": "https://arxiv.org/html/2403.08151v1#S3",
"mediaType": "application/octet-stream",
"description": "Paper detailing the BUTTER-E project and dataset."
},
{
"@type": "dcat:Distribution",
"title": "BUTTER - Empirical Deep Learning Dataset on OEDI",
"format": "HTML",
"accessURL": "https://data.openei.org/submissions/5708",
"mediaType": "text/html",
"description": "Link to the OEDI submission for the BUTTER dataset which includes a link to the original BUTTER data on AWS, data descriptions, and a tutorial Jupyter notebook for using the data."
},
{
"@type": "dcat:Distribution",
"title": "BUTTER-E GitHub ReadMe",
"format": "md",
"accessURL": "https://github.com/NREL/BUTTER-E-Empirical-analysis-of-energy-trends-in-neural-networks-supplementary-code/blob/main/Readme%20for%20Data.md",
"mediaType": "application/octet-stream",
"description": "README document describing the columns, schema, size, and format of the data contained in this submission."
},
{
"@type": "dcat:Distribution",
"title": "BUTTER-E Energy.zip",
"format": "zip",
"mediaType": "application/zip",
"description": "1-minute raw time series power data corresponding to the runs in the "BUTTER-E Metadata" resource.",
"downloadURL": "https://data.openei.org/files/5991/butter_e_energy.zip"
},
{
"@type": "dcat:Distribution",
"title": "Node Info.csv",
"format": "csv",
"mediaType": "text/csv",
"description": "Characteristics of each compute node used to generate the BUTTER-E data set.",
"downloadURL": "https://data.openei.org/files/5991/node_sinfo.csv"
},
{
"@type": "dcat:Distribution",
"title": "BUTTER-E Metadata.zip",
"format": "zip",
"mediaType": "application/zip",
"description": "Metadata concerning each training run",
"downloadURL": "https://data.openei.org/files/5991/butter_e_metadata.csv.zip"
},
{
"@type": "dcat:Distribution",
"title": "Node Power Distribution.csv",
"format": "csv",
"mediaType": "text/csv",
"description": "Power consumption quantiles for each node used to generate the BUTTER-E Dataset.",
"downloadURL": "https://data.openei.org/files/5991/node_power_dist.csv"
},
{
"@type": "dcat:Distribution",
"title": "Runs with Standardized Energy.zip",
"format": "zip",
"mediaType": "application/zip",
"description": "Power data joined to run data, including extra columns for standardized energy data as described in the paper.",
"downloadURL": "https://data.openei.org/files/5991/runs_with_standardized_energy.csv.zip"
},
{
"@type": "dcat:Distribution",
"title": "Summary by Epoch.tar",
"format": "tar",
"mediaType": "application/octet-stream",
"description": "Training losses related to the BUTTER-E dataset, re-summarized from the BUTTER dataset.",
"downloadURL": "https://data.openei.org/files/5991/summary_by_epoch.tar"
}
]
|
| DOI | 10.25984/2329316 |
| identifier | https://data.openei.org/submissions/5991 |
| issued | 2022-12-30T07:00:00Z |
| keyword |
[
"BUTTER",
"BUTTER-E",
"benchmark",
"computational science",
"deep learning",
"efficient",
"empirical deep learning",
"empirical machine learning",
"energy",
"energy consumption",
"energy efficiency",
"energy use",
"green computing",
"machine learning",
"model",
"network structure",
"neural networks",
"node-level",
"power",
"power consumption",
"training",
"training efficiency"
]
|
| landingPage | https://data.openei.org/submissions/5991 |
| license | https://creativecommons.org/licenses/by/4.0/ |
| modified | 2024-10-07T15:12:02Z |
| programCode |
[
"019:023"
]
|
| projectNumber | GO0028308 |
| projectTitle | National Renewable Energy Laboratory (NREL) Lab Directed Research and Development (LDRD) |
| publisher |
{
"name": "National Renewable Energy Laboratory",
"@type": "org:Organization"
}
|
| spatial |
"{"type":"Polygon","coordinates":[[[-180,-83],[180,-83],[180,83],[-180,83],[-180,-83]]]}"
|
| title | BUTTER-E - Energy Consumption Data for the BUTTER Empirical Deep Learning Dataset |
