Paper Title: Real-Time Hyperdimensional Quantitative Analysis using Persistent Memory
Paper Abstract: This session explores the use of Hyperdimensional Computing to enable large scale, high speed similarity searches to locate, localize and fill data tables within an on-board Persistent Memory thus enabling high speed throughput for Top K type and other similarity associated computational problems. A CXL class 3 interface mounts an FPGA based control element along with the attendant PMEM memories.
Paper Author: Gil Russell, CTO, WebFeet Research
Author Bio: Gil Russell specializes in next generation semiconductor technology for In-Memory Computing, Near Data Processing-Acceleration, Predictive Analytics, Neuromorphic Execution Units and Cognitive Intelligent Computing. He co-founded Bright Side Analytics, an IT consulting firm in July 2005 specializing in dynamic, non-volatile and persistent memory technologies. He has managed numerous new product definition activities – participating as a member of new product development engineering teams; technical lead of new product introductions; technical interface and liaison to external partnerships at Infineon Technologies, Siemens Microelectronics, NEC and Samsung.
|