Wenxuan Xu 许文轩

I am Wenxuan Xu, a Master's student in Computer Science at Dartmouth College. My current focus is AI infrastructure, especially LLM inference acceleration, serving systems, CUDA kernels, and quantization.

I previously worked on Human-Computer Interaction and VR/AR with Andrew Campbell, Hai-Ning Liang, Wolfgang Stuerzlinger, and Yuntao Wang. I eventually felt that HCI research was dead water for me, both in future prospects and in practical upside. Still, that detour gave me real research training and led to works such as LENS.

In this era, joining industry and building real systems is clearly the better choice for me. Influenced by friends who were already working on infrastructure, I started moving into LLM infra and inference acceleration. I am serious about it: I keep learning, building, and contributing to projects such as SGLang and LiteInfer. I will join Elastix AI as an AI Performance Engineer in July 2026.

我是许文轩，目前是 Dartmouth College 计算机科学硕士生。现在主要关注 AI infrastructure，尤其是 LLM inference acceleration、serving systems、CUDA kernels 和 quantization。

我之前和 Andrew Campbell、Hai-Ning Liang、Wolfgang Stuerzlinger、Yuntao Wang 做过 HCI 和 VR/AR 方向的研究。后来我逐渐感觉 HCI research area 对我来说无论前途还是钱途都是死水一潭。不过这段 detour 也给了我真正的 research training，并产出了 LENS 这样的工作。

在这个时代，我觉得加入业界、做真实系统显然是更适合我的选择。在朋友的影响下，我开始转向 LLM infra 和 inference acceleration。我对此很认真：我一直在学习、构建，也在参与 SGLang 和 LiteInfer 这样的项目。2026 年 7 月，我将加入 Elastix AI，担任 AI Performance Engineer。