NVIDIA and Google both emphasized their growing reliance on RISC-V at the annual RISC-V Summit held from October 22–24, underscoring the open ISA’s accelerating momentum across the semiconductor industry.
NVIDIA, which has already spent nine years integrating RISC-V cores into its products, delivered a 20-minute keynote led by Frans Sijstermans, Vice President at NVIDIA. The company outlined how RISC-V has become a key part of its internal architecture strategy, particularly in microcontrollers and control subsystems across its GPU platforms.
Google engineers—including Cliff Young, a co-creator of the TPU—discussed how RISC-V plays an important role in AI accelerators. Google openly highlighted the architecture’s flexibility, which enables rapid iteration and customization for advanced AI workloads. The company reaffirmed that TPUs incorporate RISC-V elements, reflecting Google’s long-term interest in the architecture.
Big Tech’s Continued Migration to RISC-V #
RISC-V’s influence is expanding as more companies seek alternatives to expensive, tightly controlled ISAs like ARM and x86.
- Qualcomm announced a major strategic commitment at last year’s summit, tied partly to its legal dispute with ARM.
- Meta publicly disclosed more details on its use of RISC-V in custom silicon.
- Apple uses RISC-V-based microcontrollers inside its M-series chips.
- Samsung recently ported TizenOS (used in its smart TVs) to RISC-V.
Across cloud and hardware providers, general support for RISC-V continues to grow, and most ecosystem players maintain active development plans.
RISC-V’s openness—royalty-free licensing, extensibility, and custom instruction options—has made it especially attractive for companies that want to reduce costs, accelerate development, or differentiate their hardware designs.
Market Momentum and Adoption Forecasts #
Analyst firm Omdia forecasts that RISC-V processor shipments could reach 17 billion units by 2030, with automotive chips accounting for ~46% of that total.
RISC-V International hopes to expand the ISA into servers and PCs, though most industry watchers agree this transition will take time due to ecosystem challenges.
Meanwhile, proprietary architectures continue to evolve:
- x86 is experiencing a small resurgence, bolstered by improved power efficiency and cooperative efforts between Intel and AMD.
- Google, Microsoft, and AWS have all developed custom ARM-based processors, further fragmenting the market.
Software Ecosystem and AI/Server Discussions #
Sessions at the summit covered updates on:
- AI accelerators
- GPU architectures
- Server-class RISC-V specifications
- Software and OS support challenges
RISC-V’s software ecosystem remains a weak point, and discussions frequently focused on improving compiler support, package compatibility, and cloud tooling—areas where x86 and ARM remain mature.
Rising Focus on Security #
Security emerged as a major theme this year. Despite a relatively quiet security landscape in the previous summit, 2024 included numerous sessions on:
- Confidential computing
- Cryptographic extensions
- Secure hardware modules
Attention intensified following the disclosure of Ghostwrite, a vulnerability affecting Alibaba’s Xuantie C910 RISC-V core that allowed unauthorized access to privileged memory. The disclosure accelerated industry discussion on formal verification and security-first extension standards.
Alibaba, a key summit sponsor, also introduced the R908, a new embedded RISC-V chip featuring enhanced security extensions.
Geopolitics Around RISC-V #
RISC-V has drawn attention from U.S. lawmakers who are concerned about China’s heavy investment in the architecture.
Security researchers warn that:
- Chinese vendors could potentially ship RISC-V chips with hidden backdoors
- RISC-V chips are harder to patch than x86 due to the absence of flexible microcode layers
Despite the growing geopolitical tension, these concerns were not part of the summit’s official agenda.
Manufacturing Challenges Kept Offstage #
While some vendors briefly mentioned RISC-V hardware roadmaps, large-scale manufacturing issues—capacity, pricing, and foundry choices—were not a major focus of public discussions. These topics remain sensitive and are typically handled privately between vendors and manufacturing partners.