AI Infra Wiki

标签: tpu

此标签下有2条笔记。

  • 2026年7月09日

    DNN Accelerator Systolic Dataflow

    • accelerator
    • dnn
    • systolic
    • dataflow
    • tpu
    • npu
    • roofline
    • dsa
  • 2026年7月07日

    AI Accelerators for Large Language Model Inference: Architecture Analysis and Scaling Strategies

    • accelerator
    • survey
    • gpu
    • tpu
    • wse
    • lpu
    • inference
    • scaling
    • expert-parallelism

Created with Quartz v4.5.1 © 2026

  • Source Wiki
  • Quartz