Chipmakers are on the speed dial for cloud giants like Google and Amazon when they need volume production of in-house designed chips. But a Google-led program is trying to open up silicon on the roadmap to combat that dominance.
Google leads a group of partners that provides open-source software tools that allow chip designers to design, verify, and test virtual versions of chips, then get the physical parts made in factories for free. The goal is to help DIYers, universities, and chip startups reduce chip design and production costs.
Contents
Advertising
More partners
In recent weeks, the Google-led effort has won more partners to boost the open-source chip effort, including maker Global Foundations, and the US Department of Defense, which last week injected $15 million. dollars in funding for SkyWater, which provides the capacity in its factories to manufacture the open source chips.
The idea of open source chips can be compared to Linux. Companies like Intel and ARM have proprietary chip designs that power computing, and companies have to pay for the privilege of using those architectures. Google’s goal is to promote open chip designs that are license-free and can be modified to meet an organization’s needs.
Last mile effort
Google’s open-source chip effort solves a last-mile problem that many chip startups can’t overcome: getting the physical parts fabricated. Google’s goal is to promote the basic design of processors and sensors used in applications such as AI and data collection, and then provide a pathway to manufacturing facilities.
According to data from Pitchbook, venture capitalists poured $9.9 billion into semiconductor companies, with most going to fabless companies – which are chip companies that create designs virtual chips on computers, but do not have their manufacturing facilities.
Some notable fabless companies include Nvidia and Apple, which send their designs to chipmakers like Taiwan Semiconductor Manufacturing Co. The fabless model has become popular with the consolidation of chip manufacturing facilities.
Until around the 2000s, many semiconductor companies like Fairchild, AMD, and Motorola had their own manufacturing facilities, but later either abandoned their assets or were acquired due to volatility in the semiconductor market and the cost maintenance of these facilities.
Effects of flea shortage
Chip shortages over the past few years have crippled many small fabless companies, including AI chipmakers, who have been unable to secure manufacturing capacity with companies like AMD, Nvidia and Apple. willing to pay a premium to secure volume manufacturing.
The chip shortages were mostly in parts such as power management and sensor circuitry, which cost pennies to dollars, and are in high demand in industries such as automotive, which is transitioning to electric cars.
Open MPW
Google’s “build your silicon” program provides guidance on how to create chips, which can then be submitted to a company-funded program to make chips at no cost.
“Google’s vision is to create a complete open source ecosystem,” said Mike Wishart, CEO of efabless, which facilitates the program called Open MPW to receive and verify chip designs. Google and its partners provide the software tools free of charge.
EFabless provides open-source EDA tools for designing and verifying chips, and SkyWater provides a process design kit to prepare the chip for manufacturing. The Open MPW program is now open and the deadline is September 12.
Focus on design
For beginners and DIY chip designers, EFabless also provides the system-on-chip called Caravel which is “a big empty user space for people to draft and simplify designs. That way they can just focus on their design,” Wishart said.
“They don’t need to wade their way into the traditional semiconductor ecosystem where there just aren’t enough people to handle all the designs…and it’s too expensive or too complex,” Wishart said.
The chips are made on SkyWater’s 130nm process, which is an older node similar to that used by Intel to make its Pentium III and 4-output chips from 2001. Some of the more advanced chips, like the A15 Bionic in the Apple’s latest smartphones, are manufactured on the most advanced 5nm process.
Smaller chipmakers usually use old process nodes due to the low manufacturing cost. These chips can be used to prototype versions of a chip before a final version is produced and released on more advanced nodes. One participant, Johannes Kepler University of Linz, used the program to create student-designed chip prototypes in the classroom.
Increased customization
The Google program makes sense from an academic and STEM perspective, and for companies developing their own chips, said Jim McGregor, principal analyst at Tirias Research.
“Pretty much everyone is trying to design AI on the platform, native or in the cloud. We’re seeing more and more chip customization,” McGregor said.
Google is developing its own closed-source chips like the Tensor Processing Unit for AI in its cloud environment, but the open-source initiative may be an effort to build goodwill with the growing number of open-source chip designers, McGregor said.
“It’s their way of connecting them to future designers and chips, much like Apple initially sells or donates PCs to academics,” he said.
Google is a member of RISC-V International, which manages the development of RISC-V, a license-free open source chip architecture. The RISC-V architecture is considered a rival to chip architectures from Intel and ARM, but is not quite there yet in terms of performance.
It’s also useful to have physical chips in their hands, but there are also many compatible chips and operating systems for parts out of the box on older process nodes.
“It’s always a question of whether they have the time or the money to develop custom chips when they’re available in the market,” McGregor said.
The New Stack is a wholly owned subsidiary of Insight Partners, an investor in the following companies mentioned in this article: Unit, Bionic.
Featured image by Alireza Hatami on Unsplash.