SRAM as Processing
Using SRAM for processing instead of just storage, known as processing-in-memory, allows data to be computed directly within memory cells. This technique reduces data movement overhead, improving energy efficiency and enabling simple logic operations within SRAM arrays.
Background
SRAM (Static Random-Access Memory) is a type of fast, volatile memory used in CPUs as cache. Normally, data is moved from SRAM to the processor's ALU (Arithmetic Logic Unit) for computation, then written back — a round trip that costs time and energy. This article discusses "Processing-Using-SRAM" (PIM), an emerging research direction where simple logic operations are performed directly inside the SRAM array, treating memory itself as the compute unit. The technique exploits the physical behavior of SRAM bitlines during read/write cycles to perform bulk bitwise operations (AND, OR, etc.) without moving data. This is part of the broader "near-memory" or "in-memory computing" movement that tries to overcome the "von Neumann bottleneck" — the speed and energy gap between CPU and memory that limits modern computing performance.