The Development of Silicon for AI: Different Design Approaches
| Autor: | Hoi-Jun Yoo, Jinmook Lee, Sungpill Choi, Kyuho Jason Lee |
|---|---|
| Rok vydání: | 2020 |
| Předmět: |
Digital electronics
Artificial neural network Machine vision business.industry Computer science media_common.quotation_subject 020208 electrical & electronic engineering 02 engineering and technology Perceptron 020202 computer hardware & architecture Neuromorphic engineering Computer architecture 0202 electrical engineering electronic engineering information engineering Electrical and Electronic Engineering Function (engineering) business Implementation media_common Block (data storage) |
| Zdroj: | IEEE Transactions on Circuits and Systems I: Regular Papers. 67:4719-4732 |
| ISSN: | 1558-0806 1549-8328 |
| DOI: | 10.1109/tcsi.2020.2996625 |
| Popis: | This paper provides a review of design approaches towards artificial intelligence (AI) System-on-Chip. AI algorithms have progressed over the past decades from perceptron-based neural network (NN) and neuro-fuzzy (NF) system to today’s deep neural network (DNN) and neuromorphic computing. Recent DNN hardware accelerators focus on energy-efficient integration of digital circuits to realize real-time DNN operation while neuromorphic processors deploy new memory technologies with analog computation for low power consumption. However, different design approaches can be applied to such processor implementation with their pros and cons. This paper reviews from the early processor designs for NN and NF in both mixed-mode and digital implementations to the recent DNN SoC designs that we have proposed for a decade. The former content deals with NN and NF processors used as a functional building block of a machine vision SoC, while the latter concentrates on integration of the whole DNN function. We also provide a discussion on the approaches, and provide perspective on future research directions. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|