Coruscant: Co-Designing GPU Kernel and Sparse Tensor Core to Advocate Unstructured Sparsity in Efficient LLM Inference | OpenReview

Coruscant: Co-Designing GPU Kernel and Sparse Tensor Core to Advocate Unstructured Sparsity in Efficient LLM Inference

Open Webpage

Donghyeon Joo, Helya Hosseini, Ramyad Hadidi, Bahar Asgari

Published: 2025, Last Modified: 09 May 2026MICRO 2025EveryoneRevisionsBibTeXCC BY-SA 4.0

External IDs:dblp:conf/micro/JooHHA25

Loading