Skip to main content
T4 Automotive and self-driving NIO Internship Round 1

NIO AI Infra internship first round (2)

Real NIO AI Infra internship first-round interview questions covering inference optimization, operator optimization, and related topics

Inference optimization Operator optimization

Coding

  1. Hand-write a CUDA kernel: reduce_sum
  2. Hand-write a CUDA kernel: conv2d
  3. Hand-write one custom kernel from the internship
  4. Implement a circular buffer in C++
  5. Bubble-sort a singly linked list in C++

Inference optimization

  1. Overall CUDA optimization work in the internship
  2. Discussion of large-model inference (vLLM, model quantization, etc.)