Linux Foundation's OPI Project Ships First Coordinated Release, Standardizing DPU and IPU Programming
OPI Abstraction v0.1.0 introduces a vendor-neutral API layer across 26 repositories and a first production deployment blueprint for DPUs and IPUs.
Overview
The Open Programmable Infrastructure (OPI) Project, a Linux Foundation initiative, has announced its first coordinated release, OPI Abstraction v0.1.0, according to a Linux Foundation press release. The release establishes a vendor-neutral API layer spanning 26 repositories, aimed at making data processing units (DPUs) and infrastructure processing units (IPUs) easier to consume, integrate, and scale across cloud and AI infrastructure deployments.
What We Know
OPI describes itself as “a community-driven initiative focused on creating a standards-based open ecosystem for next-generation architectures and frameworks based on DPU/IPU-Like” technologies, according to the Linux Foundation. The Abstraction release is intended to “break down barriers to DPU/IPU adoption, including fragmentation across proprietary SDKs, toolchains, APIs and operational models,” per the same release.
The release establishes “a vendor neutral, hardware-agnostic API layer designed to let workloads, orchestrators, and platforms program compliant DPUs without requiring hardware-specific code,” the Linux Foundation said. That layer spans 26 participating repositories “covering APIs, bridges, tooling, Kubernetes integrations, provisioning, and observability,” a figure independently confirmed by Phoronix, which reported the release “consists of a vendor-neutral API layer spanning 26 repositories.”
Alongside the API layer, OPI introduced “OPI Blueprints,” described as “a deployment-focused framework for translating open infrastructure capabilities into repeatable, production-grade solutions,” according to the Linux Foundation. The first official Blueprint, Kubernetes Network Function Offload, was developed with F5/NGINX, Intel, and Red Hat, according to Phoronix, which reported it as “a production-grade deployment pattern developed by Nginx, Intel, and Red Hat.” The Blueprint uses Red Hat OpenShift for orchestration and F5 NGINX as the application delivery controller, running on Intel’s IPU E2100 hardware, according to the OPI Project’s own announcement.
OPI also published its first white paper, titled “Accelerating the AI Era: The Critical Role of Open Programmable Infrastructure (OPI) in Scaling DPU and IPU Ecosystems,” which the Linux Foundation said “explores applications in edge computing, private cloud, data centers, high-performance networking, storage disaggregation, and zero-trust security.”
Arpit Joshipura, general manager and senior vice president of Networking, Edge and IoT at the Linux Foundation, said in the release: “AI is driving a fundamental shift in how organizations think about the infrastructure layer. DPUs and IPUs offer powerful new capabilities for offload, acceleration, security, and lifecycle management, but the ecosystem needs open APIs and repeatable deployment patterns to scale.”
Venkat Pullela, CTO of AI Infrastructure at Keysight Technologies, said DPUs are “enabling the Trifecta of CPU/GPU/DPUs to create a programmable infrastructure layer that scales,” adding that OPI is “creating a standard interface to this layer to make it easy to orchestrate across multiple nodes in a cluster.”
Joel Moses, CTO of Systems and Platforms at F5 and a member of OPI’s governing board, called the release a set of “deployment blueprints for IPU/DPU hardware,” adding that “reusable and extendable patterns are the groundwork on which enterprises build resilient architectures.”
Simon Horman, a distinguished engineer at Red Hat and chair of OPI’s Technical Steering Committee, said the Abstraction release “bridges the gap between advanced hardware acceleration and scalable, open source orchestration,” calling standardization at the infrastructure layer “essential for the future of cloud-native environments.”
OPI is scheduled to hold its second Summit on DPU/IPUs on October 15, 2026 at The Tech Interactive in San Jose, California, co-located with the OCP Global Summit, according to the Linux Foundation.
What We Don’t Know
Neither the Linux Foundation nor OPI disclosed a timeline for a v1.0 release or further Blueprints beyond the Kubernetes Network Function Offload pattern. Adoption figures — how many organizations are running the Abstraction API layer in production, or how many DPU/IPU vendors beyond Intel have committed compliant hardware — were not disclosed.
Analysis
DPUs and IPUs offload networking, storage, and security tasks from general-purpose CPUs, a capability increasingly relevant as AI infrastructure scales out across clusters. OPI’s pitch is that standardizing the software interface to this hardware — rather than leaving each vendor to ship its own SDK — reduces the integration burden for cloud operators and AI infrastructure teams. The first Blueprint’s focus on Kubernetes-native network function offload, built jointly by a networking vendor (F5/NGINX), a hardware vendor (Intel), and a platform vendor (Red Hat), signals an attempt to prove the standard works end-to-end in a real deployment stack rather than as a paper specification alone.