Skip to main content
NVIDIA
DOCA Weave Service Helm Chart
Helm Chart
NVIDIA
DOCA Weave Service Helm Chart

Helm chart for deploying the DOCA Weave Service, which provides secure multi-tenant east-west network isolation for AI factories using ASTRA, BlueField-4 DPUs, and ConnectX-9 NICs.

  • DOCA Weave Service Helm Chart

    This is the helm chart for DOCA Weave Service.

    Overview

    The DOCA Weave Service provides hardware-accelerated east-west network isolation for multi-tenant AI factories on the Vera-Rubin platform. Leveraging ASTRA, BlueField-4 DPUs securely control ConnectX-9 east-west NIC traffic independently of the host.

    Using OVS to manage ConnectX-9 E-Switches, DOCA Weave transparently applies hardware-accelerated VXLAN encapsulation and decapsulation while enforcing tenant isolation in the DPU dataplane.

    Key Features

    • East-West Network Isolation: Provides network isolation for east-west traffic on BlueField-4 DPUs used with ConnectX-9 east-west NICs.
    • Per-DPU gRPC Control Plane: Hosts a gRPC control plane on each BlueField-4 DPU above hardware flow programming, exposing the NetworkIsolation interface in this release.
    • DHCP Services: Supplies DHCP services on each ConnectX-9 NIC to provide network connectivity for tenant workers.
    • Zero-Trust DPU Enforcement: Uses ASTRA to establish a system boundary in which the host cannot directly communicate with the DPU, while the DPU securely controls the ConnectX-9 east-west data path.
    • Multi-tenant Overlay Networking: Creates isolated tenant overlay networks, allowing multiple tenants to securely share AI factory infrastructure.
    • Host-Independent Isolation: Enforces isolation autonomously on the DPU and independently of host or tenant control.
    • Hardware-Accelerated VXLAN Data Path: Uses OVS on BlueField-4 to manage ConnectX-9 E-Switches and inject hardware-accelerated VXLAN encapsulation and decapsulation flows into the NIC data path.
    • Transparent Tenant Connectivity: Transparently tunnels host traffic while retaining tenant separation in the DPU dataplane, without requiring tenant applications or workloads to manage the isolation mechanism.
    • Near Line-Rate Performance: Keeps forwarding in hardware to deliver near line-rate VXLAN encapsulation with minimal added latency and throughput impact.

    Documentation

    For information on how to use the DOCA Weave Service, visit the official repository of DPF.

    Get Help

    Enterprise Support Get access to knowledge base articles and support cases or submit a ticket.

    NVIDIA AI Enterprise Documentation

    Visit the NVIDIA AI Enterprise Documentation Hub for release documentation, deployment guides and more.

    License & EULA

    This DOCA Microservice is licensed under the NVIDIA Networking Product Agreement. By pulling and using this product, you accept the terms and conditions of this license.

    Publisher
    NVIDIA
    Latest Versionv26.8.0
    UpdatedOctober 12, 2026 UTC
    Compressed Size5.1 KB

    Modal Content

    NVIDIA uses cookies to improve your experience on our web site. We and our third-party partners also use cookies and other tools to collect and record information you provide as well as information about your interactions with our websites for performance improvement, analytics, and to assist in marketing efforts. By clicking "Accept All", you consent to our use of cookies and other tools as described in our Cookie Policy. You can manage your cookie settings by clicking on "Manage Settings." By continuing to use this site or by clicking one of the buttons below, you agree to our Terms of Service (which contains important waivers). Please see our Privacy Policy for more information on our privacy practices.