Abstract: A mixed-precision analog compute-in-memory (Mix-ACIM) is presented for mixed-precision vector-matrix multiplication (VMM). The design features an all-analog current-domain fixed-point (FxP) ...
Abstract: Analog computing-in-memory accelerators promise ultra-low-power, on-device AI by reducing data transfer and energy usage. Yet inherent device variations and high energy consumption for ...
HIPMMAPEAK is a HIP-based benchmarking tool designed to measure the peak performance of matrix multiplication operations across various data types and tensor core configurations on AMD GPUs. MMA ...
There was an error while loading. Please reload this page.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback