The latest RISC-V Summit in Santa Clara, California, brought some big announcements from major industry leaders such as Nvidia, Qualcomm, Google, and Samsung, showcasing their growing commitment to the open-source RISC-V architecture.
This event shed light on the increasing shift towards RISC-V as a viable alternative to proprietary chip architectures like Arm and x86 from Intel and AMD.
RISC-V’s flexibility, cost-efficiency, and collaborative development model are now attracting top companies that aim to enhance their AI innovations and overall performance.
Nvidia: RISC-V Powering GPU Innovation
Nvidia, which has been using RISC-V in its GPU microcontrollers for nearly ten years, delivered a compelling 20-minute keynote titled “One Architecture, Dozens of Applications, Billions of Processors.”
The session was presented by Frans Sijstermans, Nvidia’s VP of Multimedia Architecture/ASIC, who highlighted how RISC-V plays a critical role in Nvidia’s long-term plans.
The keynote explained Nvidia’s application of RISC-V for enhancing GPU design and efficiency. Sijstermans emphasized that adopting RISC-V allows the company to achieve better modularity and customization in its chips.
He pointed out that RISC-V’s open nature enables more adaptable designs tailored to specific workloads, which is crucial for the growing AI market.
Qualcomm: A Unified Approach to AI and Security
Another significant player at the summit was Qualcomm, which reaffirmed its long-term commitment to RISC-V. Qualcomm’s representatives discussed their alternative approach, called Sail, which aims to provide a single source of truth for the RISC-V Instruction Set Architecture (ISA).
This involves creating a unified base that can drive future advancements in chip design, particularly in AI and security.
Qualcomm also participated in a high-profile panel with Nvidia and other experts, focusing on the future of AI and security. This collaboration shows the increasing importance of RISC-V in addressing AI challenges and bolstering secure computing environments.
Samsung: Optimising Embedded Systems with RISC-V
Samsung used the summit to showcase how it has successfully integrated RISC-V CPUs into its embedded systems. The company’s presentation highlighted how Samsung Foundry is actively working with customers to innovate using RISC-V.
By optimizing chip and chiplet performance, Samsung aims to create more efficient, smarter systems for a wide range of applications.
Samsung’s collaboration with the RISC-V ecosystem aims to streamline development processes and reduce dependency on proprietary architectures.
This push towards openness signals Samsung’s intentions to lead in embedded systems, particularly in automotive and IoT markets, where RISC-V’s scalability can shine.
Google: RISC-V for AI Accelerators
One of the most awaited talks was given by Google DeepMind’s Cliff Young and Martin Maas. Their session explored Google’s development of new AI accelerators based on RISC-V architecture, using lessons learned from Google TPUs.
The talk was a detailed exploration of designing and deploying successful AI accelerators while tackling the hurdles encountered along the way.
Google’s use of RISC-V for its Tensor Processing Units (TPUs) underscores the architecture’s potential in high-performance computing and generative AI.
By utilizing RISC-V’s customizable nature, Google aims to create AI accelerators that are optimized for specific AI workloads, boosting efficiency and reducing costs.
The Road Ahead for RISC-V in AI and Automotive
The summit underscored how RISC-V is gaining traction in two key areas—AI and automotive sectors. As proprietary architectures face constraints in terms of scalability and licensing, more companies are gravitating towards RISC-V’s open model.
Summit speakers explored how the architecture could redefine upcoming computing trends like generative AI and high-performance computing.
Though it may take time for RISC-V to penetrate the server and PC markets, its rise in AI and automotive sectors highlights its promise.
Industry experts predict that as more companies adopt RISC-V, it could become a mainstream player challenging Intel’s x86 and Arm dominance in the near future.
