Inference Stack
The reference for how AI infrastructure actually fits together
Architectures
Catalog
Where it runs
About
Subscribe
Back to the interactive map
Silicon
TPU v6e (Trillium)
google
Google's 6th-gen TPU — higher performance-per-chip and expanded pod-scale networking vs. v5e
Official docs
View in catalog
Used by
XLA / JAX
Leads to
GCP TPU v6e Pod (Trillium)