Get Started
Home
Topics
Search
Library
Research questionHow can edge analog in-memory inference reduce energy while certifying errors from heterogeneous, imperfect accelerators?Analog in-memory accelerators save energy by computing within memory, but device faults, programming errors, noise, and converter limits distort predictions differently across chips. Activating multiple accelerators can improve reliability at substantial energy cost, while relying on one may leave its error rate uncertain.
AI
Evaluation & Benchmarks
Inference Optimization
Machine Learning
Small / On-device Models
Statistical Machine Learning
Technology
Latest papersRecent research connected to this question, newest first.RACE-AIMC: Selective Inference for Heterogeneous Analog In-Memory Accelerators at the EdgeThe source concerns selective inference across a pool of physical analog in-memory accelerators, with an offline choice of accelerator and an online accept-or-defer decision. Its evidence comes from simulations using noisy weight mappings and independent test runs, reporting mathematically bounded error rates, direct-answer coverage, and modeled energy savings rather than results from deployed hardware.research paper · Sep 2, 2026
Related questions
How can passively cooled edge devices sustain DNN inference under thermal constraints?How can complex neural-network graphs be mapped across heterogeneous SoCs to balance inference latency and throughput?How can CNN inference fit on bare-metal microcontrollers with severe memory, binary-size, and runtime limits?How can Vision Transformers avoid LayerNorm’s global reduction on edge accelerators without expensive retraining?