This article is comprehensively written by Semiconductor Industry Insights (ID: ICVIEWS)

The RISC-V architecture is making waves in AI and automotive fields.

At the recent RISC-V Summit held in Santa Clara, California, industry giants such as NVIDIA, Qualcomm, Google, and Samsung gathered to explore the potential of the RISC-V architecture. This collaboration marks a shift towards open-source computing, positioning RISC-V as a formidable alternative to the established Arm and x86 architectures.

Notably, as early as 2023, 13 companies, including Google, Intel, NVIDIA, Qualcomm, and Ali Group, launched the global RISC-V software ecosystem initiative - RISE, aimed at accelerating RISC-V's software ecosystem development and application commercialization.

NVIDIA Accelerates RISC-V GPU Shipments

For nearly a decade, NVIDIA has quietly integrated RISC-V into its GPU microcontrollers. At the summit, the company shared insights into how this architecture is transforming its product designs. In a presentation titled 'One Architecture, Dozens of Applications, Billions of Processors,' NVIDIA emphasized RISC-V's versatility. The company's focus is on how this open standard enables the development of powerful GPUs and opens doors to a wider range of applications.

So far, NVIDIA has developed at least three RISC-V microcontroller cores: NV-RISCV32 (RV32I-MU, in-order single-issue core), NV-RISCV64 (RV64I-MSU, out-of-order dual-issue core), and NV-RVV (RV32I-MU, NVRISCV32 + 1024-bit vector extension). These cores, along with potentially others, have replaced proprietary Falcon microcontroller units based on different instruction set architectures. Additionally, NVIDIA has developed over 20 custom RISC-V extensions to enhance performance, functionality, and security. Perhaps the most significant RISC-V-based component in NVIDIA's GPUs is the embedded GPU System Processor (GSP).

According to NVIDIA's official information, the first GPUs to use RISC-V-based GSPs are based on the Turing architecture. This GSP offloads kernel driver functions, reduces GPU MIMO exposure to the CPU, and manages GPU usage. Since the MCU cores are general-purpose, they can be used across all NVIDIA products. Consequently, based on one presentation, NVIDIA expects to ship around 1 billion RISC-V cores embedded in its GPUs, CPUs, SoCs, and other products by 2024, underscoring the prevalence of custom RISC-V cores in NVIDIA hardware.

Qualcomm Partners with Samsung

Qualcomm has doubled down on RISC-V and used the summit to introduce new methods for improving the RISC-V instruction set architecture. The company also participated in panel discussions with NVIDIA and other tech leaders, exploring the evolving relationship between AI development and secure computing.

Meanwhile, Samsung showcased how RISC-V CPUs have become an integral part of its embedded systems. The company emphasized its efforts to enhance chip performance and revealed how Samsung foundries help clients push boundaries with RISC-V technology.

Google DeepMind also took center stage, sharing experiences with RISC-V-based AI accelerators. Google representatives discussed successes and challenges in building Tensor Processing Units (TPUs) using the open-source architecture, emphasizing RISC-V's potential to revolutionize AI hardware.

RISC-V's Tremendous Market Potential Stems from Several Factors

Open-Source Advantage: RISC-V is an open-source instruction set architecture, allowing anyone to freely use, modify, and distribute it. This fosters innovation and customized designs with great flexibility.

Technological Progress: RISC-V is continually optimized and deeply integrated with AI technologies. For instance, Meta launched an AI inference accelerator based on RISC-V, while Qualcomm and Google collaborated on a RISC-V-based wearable chip.

Diverse Applications: RISC-V is rapidly gaining ground, with chip giants like Apple, Intel, and Qualcomm exploring its vast potential. Its applications extend to automotive, communications, robotics, ICT, power industry, and more.

Cost-Effectiveness: RISC-V's open-source nature reduces R&D and licensing costs, appealing to cost-sensitive markets.

Industry Support: Growing numbers of industry giants and startups are embracing RISC-V, encompassing chip manufacturers, software developers, and system design firms.

Market Demand: Rapid advancements in IoT, 5G, and AI drive increasing demand for high-performance, low-cost processors, positioning RISC-V as a superior choice.

RISC-V's Accelerated Adoption in AI and Automotive Fields

Omdia projects that by 2030, RISC-V architecture chip shipments will reach 17 billion, accounting for 25% of the market share, with accelerated growth in industrial and automotive sectors. This forecast signifies RISC-V's rapid market acceptance and potential to disrupt the long-standing dominance of traditional architectures.

While RISC-V chips may take time to penetrate mainstream PCs and servers, the architecture is already making waves in AI and automotive fields. Summit speakers delved into how RISC-V is reshaping the future of generative AI and high-performance computing.

With increasing corporate adoption, RISC-V appears poised to play a pivotal role in the next wave of technological innovation.

Although Intel and AMD still dominate the traditional CPU market, RISC-V's open and flexible design is steadily gaining attention. As more tech giants join forces, significant changes in the computing landscape may emerge in the coming years.

*Disclaimer: This article is written by the original author. The content reflects their personal views, and our republication is solely for sharing and discussion purposes. We do not necessarily endorse or agree with the opinions expressed. Should you have any objections, please contact us.