All-Digital SRAM-Based Full-Precision Compute-In Memory Macro in 22nm for Machine-Learning Edge Applications.
Yu-Der ChihPo-Hao LeeHidehiro FujiwaraYi-Chun ShihChia-Fu LeeRawan NaousYu-Lin ChenChieh-Pu LoCheng-Han LuHaruki MoriWei-Cheng ZhaoDar SunMahmut E. SinangilYen-Huei ChenTan-Li ChouKerem AkarvardarHung-Jen LiaoYih WangMeng-Fan ChangTsung-Yung Jonathan ChangPublished in: ISSCC (2021)
Keyphrases
- machine learning
- random access memory
- dynamic random access memory
- machine learning methods
- learning algorithm
- explanation based learning
- memory requirements
- decision trees
- natural language processing
- text mining
- power consumption
- high precision
- data mining
- memory space
- memory usage
- edge information
- precision and recall
- edge detection
- text classification
- artificial intelligence
- pattern recognition
- weighted graph
- edge detector
- learning systems
- knowledge acquisition
- associative memory
- machine learning algorithms
- supervised learning
- information extraction
- support vector machine
- past experience
- cmos technology
- feature selection
- computer vision