Leo | GPU Performance Architect & KubeGPU Maintainer

Leo

Lead Infrastructure Architect

Cloud-Native Architect

KubeGPU Maintainer

Extreme Altitude GPU Tester

Experience & Education

Speciality

A GPU performance optimization pioneer who believes "computing resources should flow as efficiently as utilities." Specializes in transforming complex cluster management into actionable solutions.

Experience

1.Senior Engineer, AWS EC2 GPU Instance Optimization Team (5 years)
2.Led a 40% failure-rate reduction project for a 1,000-GPU cluster in autonomous driving
3.Core designer of WhaleFlux’s scheduling algorithms

Education

1.MS Distributed Systems, Carnegie Mellon University
2.BS Computer Science, Tsinghua University

Posts

High Performance Computing Cluster Decoded

High Performance Computing Cluster Decoded

Leo 6 月 17, 2025

What High-Performance Computing Really Means in the AI Era

What High-Performance Computing Really Means in the AI Era

Leo 6 月 17, 2025

GPU Coroutines: Revolutionizing Task Scheduling for AI Rendering

GPU Coroutines: Revolutionizing Task Scheduling for AI Rendering

Leo 6 月 16, 2025

The Vanishing HAGS Option: Why It Disappears and Why Enterprises Shouldn’t Care

The Vanishing HAGS Option: Why It Disappears and Why Enterprises Shouldn’t Care

Leo 6 月 16, 2025

Beyond the HAGS Hype: Why Enterprise AI Demands Smarter GPU Scheduling

Beyond the HAGS Hype: Why Enterprise AI Demands Smarter GPU Scheduling

Leo 6 月 16, 2025

The Ultimate GPU Benchmark Guide: Free Tools for Gamers, Creators & AI Pros

The Ultimate GPU Benchmark Guide: Free Tools for Gamers, Creators & AI Pros

Leo 6 月 10, 2025

Enhancing LLM Inference with GPUs: Strategies for Performance and Cost Efficiency

Enhancing LLM Inference with GPUs: Strategies for Performance and Cost Efficiency

Leo 1 月 17, 2025

Enhancing LLM Inference with GPUs: Strategies for Performance and Cost Efficiency

Enhancing LLM Inference with GPUs: Strategies for Performance and Cost Efficiency

Leo 10 月 23, 2024

« Previous
1
…
8
9
10