Domain
AI Hardware
AI hardware engineers design systolic arrays, memory hierarchies, NoCs and compilers-aware datapaths for training and inference chips.
All domainsSee courses, channels, docs and software →
Daily work
- Architecture modelling and PPA tradeoffs
- RTL for compute datapaths
- Benchmark analysis
- Work with compiler teams
Required skills
- Computer architecture
- Quantisation & ML basics
- HBM/memory systems
- RTL + performance modelling
Career roadmap
- Architecture fundamentals
- ML basics
- Accelerator design patterns
- Performance modelling
Future scope
The single hottest area in semiconductors — AI chip demand drives global capex.
Learning path for AI Hardware
Beginner roadmap
- Architecture
- Deep learning basics
- Accelerator papers
- Build a small NPU
Recommended books
- Efficient Processing of Deep Neural Networks — Sze et al.
Practice projects
- Systolic array matrix multiplier in RTL
- Roofline analysis of an accelerator