Skip to main content
T3 Chips and hardware Beiji Xiongxin Round 1

Arctic Xiongxin AI Infra first round

Real Arctic Xiongxin AI Infra first-round interview questions covering operator optimization

Operator optimization

Fundamentals

  1. How familiar are you with AI compilers such as TVM and XLA?
  2. What common methods exist for thread synchronization?
  3. How do you implement the singleton pattern?
  4. What pitfalls does double-checked locking have?
  5. Besides singleton, what other design patterns do you know?
  6. Underlying implementation of malloc

CUDA and frameworks

  1. How to use NVIDIA Profiler tools
  2. How much do you know about PyTorch’s internals?
  3. Common CUDA operator optimization techniques

Other

  1. Are you familiar with common Linux commands and operations?