RF

Ruibo Fan

0009-0006-7492-2069

cs.CLcs.DCcs.LGcs.PFcs.ARcs.OS

On Valency

published · living versions

ResServe: Accurate and Efficient FP4 LLM Inference with Residue Compensation

with Shuo Rong, Jiawen Wen, Xiaowen Chu

1 version

W_5py3r4vev1 · current

Preprints & journals

8 papers in the corpus · 2023–2026

KREX: Concurrent Kernel Benchmarking on Shared GPUs via Region-Granular Exclusivity

Tianyu Feng, Haoxuan Yu, Tianyuan Wu et al.

0 citations

2609.30057v12026arXiv

ZipCCL: Efficient Lossless Data Compression of Communication Collectives for Accelerating LLM Training

Wenxiang Lin, Xinglin Pan, Ruibo Fan et al.

1 citation

2604.27844v12026arXiv

ZipServ: Fast and Memory-Efficient LLM Inference with Hardware-Aware Lossless Compression

Ruibo Fan, Xiangrui Yu, Xinglin Pan et al.

0 citations

2603.17435v12026arXiv

Dissecting Outlier Dynamics in LLM NVFP4 Pretraining

Peijie Dong, Ruibo Fan, Yuechen Tao et al.

0 citations

2602.02047v12026arXiv

Dissecting the NVIDIA Hopper Architecture through Microbenchmarking and Multiple Level Analysis

Weile Luo, Ruibo Fan, Zeyu Li et al.

1 citation

2501.12084v22025arXiv

STBLLM: Breaking the 1-Bit Barrier with Structured Binary LLMs

Peijie Dong, Lujun Li, Yuedong Zhong et al.

1 citation

2408.01803v22024arXiv

Benchmarking and Dissecting the Nvidia Hopper GPU Architecture

Weile Luo, Ruibo Fan, Zeyu Li et al.

58 citations

2402.13499v12024arXiv

Dissecting the Runtime Performance of the Training, Fine-tuning, and Inference of Large Language Models

Longteng Zhang, Xiang Liu, Zeyu Li et al.

1 citation

2311.03687v22023arXiv

Career total: 16 works. 8 are in this corpus.

Profile built from the corpus for this byline.

Author records are still filling in while the Hub is in beta. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.