Energy-Aware Deep Neural Architecture Search for Carbon-Efficient AI on Resource-Constrained Devices
Bibliographic record
Abstract
Deep learning models are being quickly put on edge and Internet of Things (IoT) devices, which have made people, worry about how much energy they use and how much carbon they release. When it comes to traditional deep neural architecture search (NAS), the main goals are to make things more accurate and faster. This article is about an Energy-Aware Deep Neural Architecture Search (EANAS) system that can help you construct deep learning models that use less energy and work better on devices with restricted resources. The proposed approach employs multiobjective optimization, simultaneously considering accuracy, latency, energy consumption, and carbon efficiency. The system makes sure that edge GPUs, ARM CPUs, and microcontrollers can all work together by putting in place hardware-aware limits. The search space offers several levels, operators, and quantization methods so that you can discover the ideal balance between speed and energy savings. To make sure the system works, tests are run on a variety of benchmarks, such as speech processing, picture recognition, and sensor-based datasets. The suggested EA-NAS models are used on real-life mobile and IoT systems, showing big energy savings in cases with ongoing inference. The framework constantly lowers energy use while keeping competitive performance, as shown by the experiments. This makes it a great choice for real-time, low-power apps.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".