Yifan Xiong

Senior Researcher

Research Focus: August 5, 2024

Microsoft Research ブログ

Research Focus: Week of August 12, 2024

8月 14, 2024

In this issue: Research Forum Ep. 4 expl…

アワード | USENIX ATC 2024

Best Paper Award at USENIX ATC 2024

July 10, 2024

Our paper titled "SuperBench: Improving …

A line graph comparing the end-to-end performance of Meta’s MoE language model using Azure NDm A100 v4 VMs with and without Tutel. The x-axis is the number of A100 (80GB) GPUs, beginning at 8 and going up to 512, and the y-axis is the throughput (K tokens/s), beginning with 0 and going up to 1,000 in intervals of 100. Tutel always achieves higher throughput than fairseq.

Microsoft Research ブログ

Tutel: An efficient mixture-of-experts implementation for large DNN model training

11月 22, 2021 | Wei Cui, Yifan Xiong, Peng Cheng, と Rafael Salas

Mixture of experts (MoE) is a deep learn…