Why it matters
Generic hardware often reaches efficiency limits as production workloads scale. Custom silicon allows development teams to optimise for specific model architectures, potentially reducing long-term compute costs and improving inference latency.
Key points
- $43 million raised in initial funding round
- Founded by industry veterans from Alphawave and Qualcomm
- Provides modular components for bespoke chip design
- Targets the growing demand for specialised AI hardware


