Computer Architecture Intern

Snap Snap · Consumer · Eindhoven - Netherlands

Computer Architecture Intern to implement, map, and simulate neural rendering algorithms on event-based Neural Processing Units (NPUs), perform benchmarking, and identify hardware/compiler improvements for efficiency.

What you'd actually do

  1. Implement, map, and simulate selected neural rendering algorithms on our NPU.
  2. Perform performance benchmarking and analysis, documenting and presenting findings to the team.
  3. Identify hardware and compiler/training improvements that make our NPU more efficient for neural rendering workloads.
  4. Develop proof-of-concept implementations to validate proposed optimizations.

Skills

Required

  • computer architecture
  • digital design
  • deep learning
  • AI algorithms
  • neural networks
  • Python
  • C++

Nice to have

  • PyTorch
  • TensorFlow
  • Unity
  • Unreal Engine
  • event-based processors
  • event-based sensors

Other signals

  • NPU development
  • neural rendering algorithms
  • performance benchmarking
  • hardware and compiler improvements