SRAM Compute-in-Memory Full-stack Architecture

Explore how VXI integrates SRAM compute-in-memory, data fabric, memory hierarchy, compiler and runtime so large models run within the power budget of robots and edge devices.

Find the technology that matters to you

Type a keyword or pick a tag to filter the six VXI Gen1 technology modules, then open the full page for each module.

  • SRAM-CIM compute path

    Weight residency and near-memory compute that cut data movement

    Weight residency and near-memory compute are the architecture direction for reducing unnecessary data movement.

  • On-chip data fabric

    Workload-aware data paths and module communication

    Coordinate data paths, traffic governance, and module communication around each workload.

  • Memory hierarchy

    Co-placed weights, activations, and cache

    Co-design weights, activations, cache, and external memory around model behavior.

  • Compiler and runtime

    From model conversion and quantization to runtime control

    Connect model conversion, quantization, scheduling, and runtime control to accelerate deployment.

  • Security and data sovereignty

    Secure boot, model protection, and local data processing

    Support enterprise deployment through secure boot, model protection, and local data processing.

  • System co-engineering

    An integration path connecting model, software, host, and memory

    Connect model, software, host, and memory needs into an executable integration path.