NVIDIA Hopper & Blackwell GPU Support Moves Closer For Open-Source Nova Driver

Key Takeaways

• Nova driver v12 rebases onto drm‑rust‑next, improving Rust code.
• Hopper and Blackwell GPUs now enabled in open‑source driver pipeline.
• Patches address Foundation Security Processor integration for GSP path.
• NVIDIA engineers actively reviewing code on kernel mailing list.
• Open‑source support could challenge proprietary NVIDIA Linux driver dominance.

Pulse Analysis

The Linux graphics stack has long relied on the community‑driven Nouveau driver, which, despite years of effort, still lags behind NVIDIA’s proprietary offering in performance and feature parity. In response, NVIDIA has launched the Nova project, a Rust‑based open‑source driver that targets the GPU System Processor (GSP) architecture used in its latest Hopper and Blackwell silicon. By exposing the same low‑level interfaces that the closed‑source stack uses, Nova promises to narrow the gap for developers who need reliable, high‑throughput GPU acceleration on Linux servers and workstations.

The twelfth patch series, released this week, marks a significant milestone. Engineer John Hubbard rebased the code against the drm‑rust‑next branch, cleaned up several Rust modules, and added support for the Foundation Security Processor (FSP), a critical component of the GSP pathway. These changes bring Hopper and Blackwell GPUs—NVIDIA’s flagship AI and HPC accelerators—into the Nova testing matrix, allowing the community to validate power management, memory handling, and compute kernels under a fully open stack. Review is now underway on the kernel mailing list, signaling a near‑final code freeze.

If Nova reaches production readiness, the implications for the data‑center market are profound. Enterprises that run AI workloads on Linux could adopt a fully open driver, reducing licensing costs and simplifying compliance with security policies that restrict proprietary binaries. Moreover, the Rust implementation offers memory safety guarantees that may lower driver‑related crash rates, a persistent pain point for large‑scale clusters. For NVIDIA, embracing open‑source could broaden its ecosystem, attract more Linux‑centric developers, and reinforce its leadership in AI hardware while still monetizing through premium software stacks.