Distributed, Parallel, and Cluster Computing

From Servers to Sites: Compositional Power Trace Generation of LLM Inference for Infrastructure Planning Grant Wilkins, Fiodar Kazhamiaka, Ram Rajagopal arxiv.org/abs/2603.18383 [𝚌𝚜.𝙳𝙲]

High-Performance Portable GPU Primitives for Arbitrary Types and Operators in Julia Emmanuel Pilliat (ENSAI) arxiv.org/abs/2603.18695 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙿𝙵]

Act While Thinking: Accelerating LLM Agents via Pattern-Aware Speculative Tool Execution Yifan Sui, Han Zhao, Rui Ma, Zhiyuan He, Hao Wang, Jianxun Li, Yuqing Yang arxiv.org/abs/2603.18897 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙰𝙸]

Literature Study on Operational Data Analytics Frameworks in Large-scale Computing Infrastructures Shekhar Suman, Xiaoyu Chu, Alexandru Iosup arxiv.org/abs/2603.19016 [𝚌𝚜.𝙳𝙲]

Why Synchronized Time is a Fiction: Daylight Saving Time, Leap Seconds, and the Guillotine Sharpened for Nothing Paul Borrill arxiv.org/abs/2603.19099 [𝚌𝚜.𝙳𝙲]

A Pipelined Collaborative Speculative Decoding Framework for Efficient Edge-Cloud LLM Inference Yida Zhang, Zhiyong Gao, Shuaibing Yue, Jie Li, Rui Wang arxiv.org/abs/2603.19133 [𝚌𝚜.𝙳𝙲]

Background and Intellectual Development: Supplementary Material for the Category Mistake Papers Paul Borrill arxiv.org/abs/2603.16907 [𝚌𝚜.𝙳𝙲 𝚙𝚑𝚢𝚜𝚒𝚌𝚜.𝚑𝚒𝚜𝚝-𝚙𝚑]

The 1/W Law: An Analytical Study of Context-Length Routing Topology and GPU Generation Gains for LLM Inference Energy Efficiency Huamin Chen, Xunzhuo Liu, Yuhan Liu, Junchen Jiang, Bowei He, Xue Liu arxiv.org/abs/2603.17280 [𝚌𝚜.𝙳𝙲]

ZipServ: Fast and Memory-Efficient LLM Inference with Hardware-Aware Lossless Compression Ruibo Fan, Xiangrui Yu, Xinglin Pan, Zeyu Li, Weile Luo, Qiang Wang, Wei Wang, Xiaowen Chu arxiv.org/abs/2603.17435 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙰𝚁 𝚌𝚜.𝙻𝙶 𝚌𝚜.𝙿𝙵]

inference-fleet-sim: A Queueing-Theory-Grounded Fleet Capacity Planner for LLM Inference Huamin Chen, Xunzhuo Liu, Yuhan Liu, Junchen Jiang, Bowei He, Xue Liu arxiv.org/abs/2603.16054 [𝚌𝚜.𝙳𝙲]

Biased Compression in Gradient Coding for Distributed Learning Chengxi Li, Ming Xiao, Mikael Skoglund arxiv.org/abs/2603.16353 [𝚌𝚜.𝙳𝙲]

An Efficient Heterogeneous Co-Design for Fine-Tuning on a Single GPU Ruijia Yang, Zeyi Wen arxiv.org/abs/2603.16428 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙰𝙸]

FleetOpt: Analytical Fleet Provisioning for LLM Inference with Compress-and-Route as Implementation Mechanism Huamin Chen, Xunzhuo Liu, Yuhan Liu, Junchen Jiang, Bowei He, Xue Liu arxiv.org/abs/2603.16514 [𝚌𝚜.𝙳𝙲]

Dataflow-Oriented Classification and Performance Analysis of GPU-Accelerated Homomorphic Encryption Ai Nozaki, Takuya Kojima, Hideki Takase, Hiroshi Nakamura arxiv.org/abs/2603.16692 [𝚌𝚜.𝙳𝙲]

Looking for (Genomic) Needles in a Haystack: Sparsity-Driven Search for Identifying Correlated Genetic Mutations in Cancer Ritvik Prabhu, Emil Vatai, Bernard Moussad, Emmanuel Jeannot, Ramu Anandakrishnan, Wu-chun Feng, Mohamed Wahib arxiv.org/abs/2603.16721 [𝚌𝚜.𝙳𝙲]

ODIN-Based CPU-GPU Architecture with Replay-Driven Simulation and Emulation Nij Dorairaj, Debabrata Chatterjee, Hong Wang, Hong Jiang, Alankar Saxena, Altug Koker, Thiam Ern Lim, Cathrane Teoh, Chuan Yin Loo, … arxiv.org/abs/2603.16812 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙰𝙸 𝚌𝚜.𝙰𝚁]

NCCL EP: Towards a Unified Expert Parallel Communication API for NCCL Amos Goldman (NVIDIA Corporation), Nimrod Boker (NVIDIA Corporation), Maayan Sheraizin (NVIDIA Corporation), … arxiv.org/abs/2603.13606 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙰𝙸 𝚌𝚜.𝙰𝚁 𝚌𝚜.𝙻𝙶]

Audo-Sight: AI-driven Ambient Perception Across Edge-Cloud for Blind and Low Vision Users Jacob Bradshaw, Mohsen Riahi Alam, Bhanuja Ainary, Minseo Kim, Mohsen Amini Salehi arxiv.org/abs/2603.13668 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙲𝚈 𝚌𝚜.𝙷𝙲]

Grassroots Bonds: A Grassroots Foundation for Market Liquidity Ehud Shapiro arxiv.org/abs/2603.13671 [𝚌𝚜.𝙳𝙲 𝚌𝚜.𝙼𝙰 𝚌𝚜.𝙿𝙻]

The Markovianity of Time: The Category Mistake in Open Quantum Systems Paul Borrill arxiv.org/abs/2603.13738 [𝚌𝚜.𝙳𝙲]