T0 Big tech ByteDance Round 1
ByteDance AI Infra first round (2)
ByteDance AI Infra first-round interview questions covering AI Infra
Operator optimization
- What common optimization strategies exist for CUDA programming?
- What optimization methods apply to convolution?
- How does GPU shared memory work, and how do you use it?
Fundamentals
- How does C’s malloc differ from C++‘s new?
- What are the four C++ cast operators, and when is each used?
- What is the difference between deep copy and shallow copy?
- What types of C++ smart pointers exist, and how are they implemented?
- How do you prevent and diagnose memory leaks?
- What common Vim commands do you use?
- How do you use GDB for debugging?
- Do you have hands-on experience deploying AI models?
Project experience
- Introduce your project experience.
Coding
- Find the length of the shortest connecting path in a graph.