Anand Raghunathan
4 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
SPARQLe is a hardware-software co-design framework that exploits the inherent sub-precision sparsity of LLM activations to reduce memory traffic and enable efficient computation on lower-bit datapaths, significantly accelerating inference.
SPARQLe is a hardware-software co-design framework that exploits the inherent sub-precision sparsity of LLM activations to reduce memory traffic and enable efficient computation on lower-bit datapaths, significantly accelerating inference.
The paper proposes PowerScope, an ML-based framework for intra-cycle power estimation, achieving high accuracy and fast speed compared to commercial tools.
This paper presents TADA, a memory technology-aware DNN accelerator performance and energy estimator.