Experts at the Table — Part 1: Semiconductor Engineering sat down to talk about AI and the latest issues in SRAM with Tony Chan Carusone, CTO at Alphawave Semi; Steve Roddy, chief marketing officer at ...

“AI chips commonly employ SRAM memory as buffers for their reliability and speed, which contribute to high performance. However, SRAM is expensive and demands significant area and energy consumption.

Data refresh and cell-layout issues must be addressed to optimally implement this space-saving alternative to SRAM technology. As the system-on-a-chip (SoC) era marches forward, there's a pressing ...