| 일 | 월 | 화 | 수 | 목 | 금 | 토 |
|---|---|---|---|---|---|---|
| 1 | 2 | |||||
| 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| 10 | 11 | 12 | 13 | 14 | 15 | 16 |
| 17 | 18 | 19 | 20 | 21 | 22 | 23 |
| 24 | 25 | 26 | 27 | 28 | 29 | 30 |
| 31 |
- 코이딥
- Battery Management System
- Deep learning
- Battery SOH
- 코드로 이해하는 딥러닝
- 배터리 진단
- 칼만필터
- AzureML
- Azure
- 딥러닝
- Battery AI
- 배터리 딥러닝
- eis
- 배터리 AI
- 리튬배터리
- 머신러닝
- tensorflow
- Battery Deep Learning
- 머신러닝 코드
- 텐서플로우
- state of health
- Incremental Capacity Analysis
- bms
- 배터리 열화
- 배터리 모델링
- Battery modeling
- 배터리 EIS
- 배터리 연구
- Machine Learning
- 딥러닝 코드
- Today
- Total
목록Model of Experts (3)
Engineering insight
FlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement for Mixture-of-Experts Inference on Edge DevicesarXiv:2601.17063v1 (2026) · Sangyeob Kim et al. · HTML: https://arxiv.org/html/2601.17063v1 FlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement for Mixture-of-Experts Inference on Edge DevicesFlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement ..
FlashMoE: Fast Distributed MoE in a Single KernelNeurIPS 2025(표기: To appear) · Omotayo A. Yamenja, Divyansh Sharma, Ryan Prout, Yiran Chen · arXiv:2506.04667v3 · 링크: https://arxiv.org/abs/2506.04667v3 · 5줄 요약이 논문은 분산 Mixture-of-Experts(MoE) 추론의 핵심 병목이 expert 수학 자체보다 CPU 주도 실행 구조, 동기식 All-to-All, 잦은 커널 런치에 있다는 점을 겨냥합니다.핵심 아이디어는 gate→dispatch→expert FFN→combine→inter-GPU communication 전체를 하나의 pe..
iMOE: prediction of second-life battery degradation trajectory using interpretable mixture of expertsNature Communications (2026) · Xinghao Huang et al. · 원문: https://www.nature.com/articles/s41467-026-69369-1 한 줄 핵심이 논문은 retired EV battery의 second-life 사용에서 가장 어려운 문제인 미래 degradation trajectory 예측을, 긴 사용 이력 없이 현장 접근 가능한 부분 신호 + interpretable mixture-of-experts로 푸는 battery-AI 논문입니다. 핵심은 단순 현재 SOH..
