IBM unveils analog AI chip for deep learning inference
IBM Research introduced a mixed-signal analog AI chip for running a variety of deep neural network (DNN) inference tasks.
The device has been tested to be as adept at computer vision AI tasks as digital counterparts, while being considerably more energy efficient.
The chip was fabricated in IBM’s Albany NanoTech Complex, and is composed of 64 analog in-memory compute cores (or tiles), each of which contains 256-by-256 crossbar array of synaptic unit cells. Compact, time-based analog-to-digital converters are integrated in each tile to transition between the analog and digital worlds. Each tile is also integrated with lightweight digital processing units that perform simple nonlinear neuronal activation functions and scaling operations.
The chip also has digital communication pathways at the chip interconnects of all the tiles and the global digital processing unit.
Share this article
Help others discover this reporting.
Editorial responsibility
Jim Carroll
Editor & Publisher
Every article published by Converge Digest is researched, curated, fact-checked and editorially reviewed by Jim Carroll, Editor & Publisher. AI-assisted drafting may be used to accelerate production, but all content is reviewed, refined and approved prior to publication.
You just read about IBM. One of the companies whose work it now carries:
Cleversafe built Large-scale object storage using information-dispersal and erasure-coding techniques. What became of it?
Cleversafe brought Large-scale object storage using information-dispersal and erasure-coding techniques to Storage. That work was distinctive enough to lead to its acquisition by IBM for US$1.3B. The standalone name now passes into the archive — but the team and technology continue inside IBM today. One of 430 names here whose work outlived the company.
You just read about Semiconductors. A technology from the archive that helped get us here:
GaAs digital logic served 1980–1990. What ultimately replaced it?
GaAs digital logic gallium arsenide's higher electron mobility promised faster chips than silicon and powered the ill-fated Cray-3. Its job passed to Silicon CMOS scaling — but the problem it solved, and the ideas it proved, live on in what replaced it. The standard rests here in the archive. One of 198 technologies here that opened a door and then held it for the next.
