OpenGeMM couples a parameterized GeMM accelerator with a lightweight RISC-V core and a multi-banked scratchpad to keep accelerator utilization above 81% across DNN workloads in RTL simulation.
Aˆ 3: Accelerating attention mechanisms in neural networks with approximation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AR 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
OpenGeMM: A High-Utilization GeMM Accelerator Generator with Lightweight RISC-V Control and Tight Memory Coupling
OpenGeMM couples a parameterized GeMM accelerator with a lightweight RISC-V core and a multi-banked scratchpad to keep accelerator utilization above 81% across DNN workloads in RTL simulation.