114 citations
- Institut Polytechnique de ParisFR3 papers
- Télécom SudParisFR3 papers
- Intel (United Kingdom)GB2 papers
- The University of Texas at AustinUS2 papers
- Universitat Politècnica de CatalunyaES2 papers
- University of California San DiegoUS2 papers
- University of RochesterUS2 papers
- Washington University in St. LouisUS2 papers
- Armada (Poland)PL1 paper
- ARM (United Kingdom)GB1 paper
- Artificial Intelligence in Medicine (Canada)CA1 paper
- Barcelona Supercomputing CenterES1 paper
Showing cs.ARShow all
2 papers · 1 filter
cs.AR2025
Ember: A Compiler for Efficient Embedding Operations on Decoupled Access-Execute Architectures
Marco Siracusa, Olivia Hsu, Victor Soria-Pardos +8
Irregular embedding lookups are a critical bottleneck in recommender models, sparse large language models, and graph learning models. In this paper, we first demonstrate that, by o…
cs.AR2021
Virtual-Link: A Scalable Multi-Producer, Multi-Consumer Message Queue Architecture for Cross-Core Communication
Qinzhe Wu, Jonathan Beard, Ashen Ekanayake +2
Cross-core communication is increasingly a bottleneck as the number of processing elements increase per system-on-chip. Typical hardware solutions to cross-core communication are o…