Skip to main content
T2 AI unicorns Megvii Internship Round 1

Megvii AI Infra internship round 1

Megvii AI Infra internship round-1 interview questions covering inference optimization and operator optimization

Inference optimization Operator optimization

Fundamentals

  1. Introduce the core mechanisms and traits of the vLLM inference framework.

Project experience

  1. Introduce internship work, with focus on CUDA operator optimization practice.
  2. Technical background and main work of the project.
  3. What CUDA operator optimizations were done in the project? Concrete techniques?

Coding

  1. Find the k-th node from the end of a linked list.