Hardware

AMD Instinct MI355X

AMD's Instinct accelerator positioned against Nvidia's Blackwell, on which Wafer served GLM-5.2 at 2,626 tokens per second per node for less than half Blackwell's cost.

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

The MI355X is AMD’s Instinct accelerator positioned against Nvidia’s Blackwell (product page); AMD’s page lists a CDNA4 architecture on TSMC 3nm and 6nm FinFET lithography. In this corpus it appears once: Wafer served GLM-5.2 on it at less than half the cost of the B200 generation, with the software gap that long kept AMD out of frontier serving said to be closing because AI agents now write the kernels.

Development timeline

Newest reporting dates first. Reporting dates may differ from event dates.

    Knowledge graphopen in browser →

    Developed by
    Fabricated by

    Cite or correct this record

    Suggest a correction — opens a GitHub draft for you to review and submit.