//php echo do_shortcode(‘[responsivevoice_button voice=”US English Male” buttontext=”Listen to Post”]’) ?>
RISC-V represents an existential threat to Arm, and a new industry consortium plans to increase that threat of extinction by accelerating the development of open-source software for the RISC-V architecture. Members of the consortium this week announced the formation of the RISC-V Software Ecosystem (RISE) Project to do so.
The RISC-V Instruction Set Architecture (ISA) is a microprocessor phenomenon that started as an academic project at the University of California, Berkeley (UC Berkeley) in 2010. Since then, the open-source RISC-V ISA has developed “the big ’mo”—momentum.
Back in the 1990s, Arm processors had the big ’mo, burrowing into the heart of nearly every mobile phone. In the 2000s, they wiped out most of the proprietary microcontroller (MCU) architectures scattered over the embedded landscape.
Today, Arm is increasingly in danger of losing leadership in those arenas, and everywhere else that Arm plays. While Arm retains a long list of licensees today, earned by decades of hard work, the RISE Project could help RISC-V surge in the processor-core ranks even more quickly by bringing some order and adult supervision to the ongoing development of RISC-V open-source software.
To get a sense of the industry backing for RISE and an ascendant RISC-V ecosystem, look at the initial roster for the consortium’s governing board: Andes, Google, Intel, Imagination Technologies, MediaTek, Nvidia, Qualcomm Technologies, Red Hat, Rivos, Samsung, SiFive, T-Head and Ventana.
If this was the wild west in the U.S., that list would represent the Clanton Gang of Arizona, all gunning for one thing.
The target could be interpreted as either the elevation of the RISC-V ISA into the one true processor architecture for the non-x86 world, or the demise of Arm’s hold on that world.
The two outcomes are essentially equivalent. RISC-V has also gained traction in China, where the open ISA is not subject to political constraints or embargos.
Creation of the RISE Project represents a recognition within the RISC-V community that the essential ecosystem for the RISC-V ISA is not developing as fast or as efficiently as it should, as clearly voiced by the RISE Project’s Chair, Amber Huffman: “The RISE Project brings together leaders with a shared sense of urgency to accelerate the RISC-V software ecosystem readiness in collaboration with RISC-V International,” she said in prepared remarks.
It is that sense of urgency, more than anything else, that has been lacking in the worldwide RISC-V movement.
The situation is different for RISC-V hardware development. The RISC-V Foundation, launched in January 2015 with 36 members, helped the growing RISC-V movement to identify and codify the basic instruction set and instruction extensions. RISC-V International, founded two years later, continued the foundation’s work, and the ISA has grown steadily ever since. RISC-V International currently has nearly 3,200 members in more than 70 countries.
Back in 2018, during a presentation to the annual dinner meeting of the IEEE-CNSV (IEEE Consultants’ Network of Silicon Valley), David Patterson, who was director of the Parallel Computing Lab at UC Berkeley, concluded his presentation by saying, “That’s my simple goal for RISC-V: world domination.”
According to Jim McGregor, my Tirias Research colleague, even the manufacturing risk associated with RISC-V has been addressed.
With many MCU and system-on-chip (SoC) designs using RISC-V cores in production, there is no longer a significant risk of the costly product delays and product respins often associated with a new processor architecture. Multiple vendors now offer RISC-V processor cores and chip designers can even roll their own; something that is not possible with either Arm or x86 processors. EDA design tool vendors and major semiconductor foundries all support RISC-V cores.
And yet, the RISC-V movement has suffered from the same malaise that has afflicted countless earlier processor architectures for half a century: the hardware is willing, but the software is weak.
One of the aspects of microprocessor leadership that Arm has gotten right over the past three decades is the construction of a massive ecosystem. Arm is rightfully proud of its ecosystem, and the company has estimated that the ecosystem cuts project development costs for Arm-based projects by 50%. That is a believable number.
With a large ecosystem, chances are good that someone, somewhere, has already trod the technical paths a development team needs to travel to reach a project destination successfully. In short, a large ecosystem helps reduce project risks, and Arm’s ecosystem is quite large.
Those working on the RISE Project are committed to contributing to it financially and must provide engineering talent (the equivalent of two full-time engineers) to address specific software deliverables prioritized by the RISE Technical Steering Committee.
The RISE Project’s purpose is to encourage the development of a robust software ecosystem specifically for application processors based on the RISC-V ISA. That includes software development tools, virtualization support, language runtimes, Linux distribution integration, and system firmware by working with existing open-source communities and employing open-source best practices. These goals align well with the open-source essence of the RISC-V ISA.
Notably, Google is on the RISE Project’s governing board, and the company has committed to porting the Android operating system (OS) to the RISC-V ISA. Of course, Android is the OS used in nearly all mobile phone designs, with the exception of Apple, and the Android OS has become increasingly popular as an embedded OS, as well.
Lars Bergstrom, director of engineering on Android at Google, confirmed the company’s support for the port of the Android OS in the press release announcing the RISE Project. In addition, RISE Project’s Huffman is a principal engineer at Google.
None of this means that Arm is about to become extinct.
Arm has a tremendous head start in the market where it plays best: mobile phones and the embedded market. Arm is very well entrenched in these markets, and many segments within these markets are extremely slow to transition to new product designs—even with an existing architecture.
However, the RISC-V community’s dedication to world domination, as spoken aloud by Patterson in 2018, calls to mind baseball great Satchell Paige’s motto, as published in “Collier’s” magazine in 1953: “Don’t look back. Something might be gaining on you.”
Although it may take years for RISC-V to put a dent in Arm’s substantial processor-core market share, Arm is clearly seeing RISC-V in its rear-view mirror at a time when the company is trying to increase licensing fees and boost royalty rates in preparation for an IPO.
Meanwhile, the RISE Project promises to give the RISC-V ecosystem a sorely needed boost.