Compute with a purpose
Explore a proposed accelerator building block through workloads, data movement and scheduling.
Make room
for the next compute.
An independent concept exploring AI accelerator architecture and developer-first systems.
Bring memory closer
Memory placement is an architecture question. Examine the tradeoffs before quoting throughput.
Make the system legible
A developer view should explain how work reaches the hardware and where a bottleneck begins.
Explore before you assume.
The public demonstration makes one part of the experience tangible. It uses example data in your browser and does not connect to external services.
Open the interactive demonstration ↗