Predictive first-principles simulations for co-designing next-generation energy-efficient AI systems
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914380769132544 |
|---|---|
| author | Mamaluy, Denis Udoy, Md Rahatul Islam Mendez, Juan P. Feinberg, Ben Pan, Wei Aziz, Ahmedullah |
| author_facet | Mamaluy, Denis Udoy, Md Rahatul Islam Mendez, Juan P. Feinberg, Ben Pan, Wei Aziz, Ahmedullah |
| contents | In modern generative-AI workloads, matrix-vector/matrix-matrix multiplications (\emph{MatMul}) dominate the compute and energy cost. Achieving dramatic reductions in energy per token therefore requires a novel, specialized hardware that is co-designed across materials, devices, interconnects, circuits, and architectures rather than optimized at any single layer in isolation. In this \emph{Perspectives} article, we argue that \emph{predictive} (first-principles, fitting-parameter-free) device and interconnect simulations can close the loop between nanoscale physics and workload-level metrics, enabling the identification of device/interconnect operating regimes that plausibly support \emph{orders-of-magnitude} improvements in energy efficiency of AI accelerators. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_08995 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Predictive first-principles simulations for co-designing next-generation energy-efficient AI systems Mamaluy, Denis Udoy, Md Rahatul Islam Mendez, Juan P. Feinberg, Ben Pan, Wei Aziz, Ahmedullah Other Condensed Matter In modern generative-AI workloads, matrix-vector/matrix-matrix multiplications (\emph{MatMul}) dominate the compute and energy cost. Achieving dramatic reductions in energy per token therefore requires a novel, specialized hardware that is co-designed across materials, devices, interconnects, circuits, and architectures rather than optimized at any single layer in isolation. In this \emph{Perspectives} article, we argue that \emph{predictive} (first-principles, fitting-parameter-free) device and interconnect simulations can close the loop between nanoscale physics and workload-level metrics, enabling the identification of device/interconnect operating regimes that plausibly support \emph{orders-of-magnitude} improvements in energy efficiency of AI accelerators. |
| title | Predictive first-principles simulations for co-designing next-generation energy-efficient AI systems |
| topic | Other Condensed Matter |
| url | https://arxiv.org/abs/2603.08995 |