A model used by the architecture team to predict the chip's performance on target workloads.
Model Work product in the SoC and full-chip lifecycle and the AI accelerator lifecycle
It predicts how the chip will perform on target workloads, so architecture decisions and product plans rest on numbers rather than intuition. Performance claims in the product plan are otherwise unsupported, and shortfalls are found only after RTL or silicon.
The full lifecycle detail, with its charts, flows, tables and every section, is for registered members. Registration is free and takes a minute. You come back to this page afterwards.