Announcing the 2025 AI & RISC-V Gemini Credit Recipients

Abstract

RISC-V International announced the recipients of the 2025 AI & RISC-V Gemini Credit grants, a program designed to leverage AI models to significantly accelerate the RISC-V software ecosystem. Recipients receive up to $50,000 USD in Gemini compute credits to pursue projects focusing on AI-driven porting methods, essential tooling development, and critical porting for server use cases. This initiative supports five key research groups from global institutions collaborating with the RISE Project to drive innovation and growth.

Report

AI & RISC-V Gemini Credit Recipients Report

Key Highlights

• Grant Announcement: RISC-V International announced the winners of the 2025 AI & RISC-V Gemini Credit program.
• Program Goal: To use AI, specifically leveraging powerful models like Gemini, to drastically speed up and improve the RISC-V software ecosystem, focusing primarily on porting efforts.
• Financial Support: Recipients were awarded up to $50,000 USD in Gemini compute credits, provided by Google/DeepMind.
• Key Collaborators: The research will involve collaboration with RISC-V International and the RISC-V Software Ecosystem Project (RISE).
• Recipients: Five research teams/individuals were selected from institutions including Iowa State University, Barcelona Supercomputing Center, TU Wien, Mohamed bin Zayed University of Artificial Intelligence, and the University of California, Berkeley.

Technical Details

• AI Architecture: The core technical resource is the provision of Gemini compute credits, enabling grantees to utilize advanced AI/LLM capabilities.
• Innovative Research Focus: Projects are centered on creating new AI-driven methodologies, including developing algorithms, fine-tuned models, and agentic approaches to automate software porting processes.
• Tooling Development: Grantees are tasked with building generalizable infrastructure and tools critical for acceleration, such as automated test suite generators, specialized simulators, and formal verification systems.
• Critical Porting Target: A specific objective for the funding is the direct porting of key software, with a strong emphasis on achieving functional server use cases for the RISC-V architecture.

Implications

• Accelerated Software Maturity: By harnessing state-of-the-art AI, RISC-V aims to overcome the traditional bottleneck of manually porting large software libraries, rapidly closing the gap with established ISAs.
• Strategic Focus on Servers: Prioritizing critical software porting for server use cases signifies a serious strategic push towards making RISC-V a viable, high-performance option for data centers and enterprise computing.
• Validation of AI in ISA Development: The partnership and investment validate the crucial role that large language models and generative AI tools will play as fundamental infrastructure for future hardware and Instruction Set Architecture development.
• Strengthening Global Ecosystem: The selection of diverse global academic institutions ensures that cutting-edge research in AI-driven automation is integrated directly into the RISC-V ecosystem, fostering innovation and collaborative growth.