How China is Building an Open National Chip Plan Around RISC-V – HPCwire
Since 1987 – Covering the Fastest Computers in the World and the People Who Run Them
Since 1987 – Covering the Fastest Computers in the World and the People Who Run Them
July 19, 2023
It is becoming clearer that China’s plan to cut reliance on Western chip technology revolves around homegrown chips built using the open RISC-V architecture, which is also gaining popularity in the Europe and U.S.
China has for years has talked about moving its chip strategy in the direction of RISC-V, which is a free-to-license API blueprint to build chips, but no serious action was taken.
Previous efforts to make sovereign chips around alternative architectures went nowhere. But this year the China government has finally shown seriousness about funding RISC-V initiatives.
The Chinese government’s sudden interest in RISC-V came overnight, and that was a surprise to academic Yungang Bao, who is the deputy director of information and communications technologies at the Chinese Academy of Sciences (CAS).
“This year, the Ministry of Science and Technology and the National Science Foundation China are already inviting proposals for RISC-V related research. This is quite a substantial change,” Yungang said during a presentation at the RISC-V Summit held last month in Barcelona.
In 2012, China’s Ministry of Industry and Information Technology proposed to merge all kinds of chip architecture blueprints being used in the country into a unified design. Those architectures included x86, MIPS, PowerPC, Alpha, and SPARC. 
But it became difficult to reach a consensus on how to merge these instruction-set architectures, Yungang said.
“RISC-V provides an excellent answer,” Yungang said.
RISC-V, which is free and open-source, provides a faster and cheaper alternative to design and manufacture chips without relying on Western technologies.
“Over the past years most RISC-V related work was done, kind of, bottom-up. Today the Minister of Science and Technology is … already paying attention to RISC-V. More universities use RISC-V for teaching, for example. There is already … a textbook based on RISC-V now,” Yungang said.
In some areas, the U.S. has restricted semiconductors in trade and policy and tightened certain Western AI chip and CPU exports to Chinese organizations. That has forced China to look internally for technologies to develop chips, and RISC-V is emerging as a top option.
RISC-V research and development started in 2015 mostly at the academic and startup levels but is now rapidly maturing. RISC-V is considered an open technology, much like Linux. The base architecture can be licensed for free and then can be modified by Chinese companies to meet their needs. The most common commercial alternatives are either to use the widespread x86 architecture, that can only be purchased in manufactured form, or to purchase a licensed design for ARM architectures.
“Instructions should be free. That’s a totally different mindset.” Yungang said, adding that “many Chinese companies adopt this mindset widely.”
RISC-V development is managed by RISC-V International, which is a neutral organization and has declared the chip architecture as border-less technology.
China has had many setbacks in building its homegrown chip ecosystem. The Chinese government has been defrauded to invest billions of dollars in non-existent chip factors, and internally developed chip architectures such as Loongson, which is reportedly a mishmash of RISC-V and MIPS architectures and has not been widely adopted.
China’s efforts mirror attempts in Europe to create a set of sovereign chips based on the RISC-V and reduce the reliance on companies like Intel and ARM for chips. The EU-funded European Processor Initiative is designing RISC-V chips for AI, supercomputers, automobiles, and other electronics.
CAS is playing the long game to make RISC-V the default architecture in the China market. The grassroots effort is focused on drawing academic and startup interest in RISC-V. For example, CAS’s goal is to introduce chip design based on RISC-V early in the education process.
In 2019, the Chinese Academy of Sciences launched a countrywide initiative to promote RISC-V across China. The U.S. has placed the Chinese Academy of Sciences on an entity list due to its affiliation with the Chinese government.
Chinese organizations also established the China RISC-V Alliance in 2018 to build out a full open-source chip ecosystem by 2030. About 70 Chinese companies are now members of RISC-V International, only behind the EU, which has 87 members, and the U.S., which has 77 members.
The top 10 RISC-V startups in China have raised close to $1.18 billion in funding from venture capital firms, according to a slide in Yungang’s presentation.
CAS is working with top Chinese companies, including Alibaba, Tencent, and ZTE, to co-develop XiangShan-v3, which will match the performance of ARM’s Neoverse-N2 server CPU design, which ARM announced in 2021.
But Yungang acknowledged that there are performance gaps in some of those chips compared to the latest chips released by ARM. The most recent XiangShan-v2, for example, was slightly faster than ARM’s Cortex-A76 smartphone CPU, which was announced in 2018. But the performance gap is closing as the RISC-V development and verification matures.
Advanced GPUs and DPUs made using the 7-nm process will also be lined up for use with XiangShan CPU this year, according to a slide during Yungang’s presentation.
Chinese companies have already released many RISC-V boards. In 2023, Sophon released a 64-core RISC-V CPU, and other companies that include StarFive and Allwinner have RISC-V CPU designs and boards that can be purchased through Chinese retail sites.
The RISC-V ecosystem in China is built on the open-source ethos, with the community working together to improve designs and software. Yungang’s long-term goal is to create a platform with full chip development platform with open-source chip designs, electronic design automation, and verification tools.
“The open-source chip ecosystem can lower the barrier of chip development by saving time to market and the cost of IP, EDA tools, and engineers.” Yungang said.
The platform will also automate chip design, so processors for various applications can be released quickly.
“Over ninety percent of the code provided by the platform can be easily reused. And then the customized design only needs … less than 10% of code.” Yungang said.
Another initiative called “One Chip One Student” (OSOC) teaches undergraduate students how to design RISC-V chips.
In 2019, five undergraduate students in the OSOC program could design a RISC-V processor in four months. The Linux-compatible chip was taped out on the 110-nm process.
The OSOC has been a considerable success since, with more than 4,000 undergraduate students participating in the program until this year. About 300 universities are participating in the initiative.
U.S. semiconductor companies are facing an acute shortage of talent in the fast-growing semiconductor sector, due to the high education costs and low interest in engineering courses.
A lot of RISC-V development in the U.S. is in the hands of the private sector and has faced many setbacks. Most notably, cash-strapped Intel pulled close to $1 billion in funding it had reserved to promote RISC-V development.
Chinese institutions are currently developing open-source EDA tools to design RISC-V chips. Three tape-outs have been completed on the 110-nm and 28-nm nodes, but it is not viable for commercial use.
“The performance, the PPA (power-performance-area) is still lower than the commercial EDA tools, but it works. And then next we will let the students use the open EDA tools to build their open-source chips,” Yungang said.
To some degree the U.S. government restrictions have stifled China’s semiconductor ecosystem by blocking access to some EDA tools for advanced chips. But Yungang repeated that China’s RISC-V efforts are borderless and can be adopted by anyone in the world.
“Everyone can access the XiangShan project from GitHub. It already has … more than 400 forks,” Yungang said.
More Off The Wire
Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!
May 20, 2024
UAE-based Core42 is building an AI supercomputer with 172 million cores which will become operational later this year. The system, Condor Galaxy 3, was announced earlier this year and will have 192 nodes with Cerebras Read more…
May 17, 2024
On Tuesday May 14th, Google announced its sixth-generation TPU (tensor processing unit) called Trillium. The chip, essentially a TPU v6, is the company’s latest weapon in the AI battle with GPU maker Nvidia and clou Read more…
May 16, 2024
The 2024 ISC 2024 competition welcomed 19 virtual (remote) and eight in-person teams. The in-person teams participated in the conference venue and, while the virtual teams competed using the Bridges-2 supercomputers at t Read more…
May 16, 2024
Nvidia’s new Grace Hopper Superchip (GH200) processor has landed in nine new worldwide systems. The GH200 is a recently announced chip from Nvidia that eliminates the PCI bus from the CPU/GPU communications pathway. Read more…
May 16, 2024
What an interesting panel, Quantum Advantage — Where are We and What is Needed? While the panelists looked slightly weary — their’s was, after all, one of the last panels at ISC 2024 — the discussion was fascinat Read more…
May 15, 2024
AI is one of the most transformative and valuable scientific tools ever developed. By harnessing vast amounts of data and computational power, AI systems can uncover patterns, generate insights, and make predictions that Read more…
May 17, 2024
On Tuesday May 14th, Google announced its sixth-generation TPU (tensor processing unit) called Trillium. The chip, essentially a TPU v6, is the company’s l Read more…
May 16, 2024
What an interesting panel, Quantum Advantage — Where are We and What is Needed? While the panelists looked slightly weary — their’s was, after all, one of Read more…
May 15, 2024
AI is one of the most transformative and valuable scientific tools ever developed. By harnessing vast amounts of data and computational power, AI systems can un Read more…
May 15, 2024
The makers of the Aurora supercomputer, which is housed at the Argonne National Laboratory, gave some reasons why the system didn’t make the top spot on the Top Read more…
May 15, 2024
Some scientific computing applications cannot sacrifice accuracy and will always require high-precision computing. Therefore, conventional high-performance c Read more…
May 14, 2024
The Linux Foundation, the nonprofit organization enabling mass innovation through open source, is excited to announce the launch of the High-Performance Softw Read more…
May 13, 2024
First, the top line: the overall HPC market was flat in 2023 at roughly $37 billion, bogged down by supply chain issues and slowed acceptance of some larger sys Read more…
May 13, 2024
The 63rd installment of the TOP500 list is available today in coordination with the kickoff of ISC 2024 in Hamburg, Germany. Once again, the Frontier system at Read more…
February 8, 2024
Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…
August 17, 2023
The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…
October 30, 2023
With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…
December 11, 2023
Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…
January 25, 2024
In under two minutes, Meta’s CEO, Mark Zuckerberg, laid out the company’s AI plans, which included a plan to build an artificial intelligence system with the eq Read more…
October 5, 2023
When discussing GenAI, the term “GPU” almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word “GPU” is assumed to mean “Nvidia” products. (As an aside, the popular Nvidia hardware used in GenAI are not technically… Read more…
March 18, 2024
Nvidia’s latest and fastest GPU, codenamed Blackwell, is here and will underpin the company’s AI plans this year. The chip offers performance improvements from Read more…
May 15, 2024
The makers of the Aurora supercomputer, which is housed at the Argonne National Laboratory, gave some reasons why the system didn’t make the top spot on the Top Read more…
January 30, 2024
Early quantum computing pioneer D-Wave again asserted – that at least for D-Wave – the commercial quantum era has begun. Speaking at its first in-person Ana Read more…
December 7, 2023
AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…
February 1, 2024
The immediate effect of the GenAI GPU Squeeze was to reduce availability, either direct purchase or cloud access, increase cost, and push demand through the roof. A secondary issue has been developing over the last several years. Even though your organization secured several racks… Read more…
August 8, 2023
Intel is planning to onboard a new version of the Falcon Shores chip in 2026, which is code-named Falcon Shores 2. The new product was announced by CEO Pat Gel Read more…
May 7, 2024
We have all thought about it. No one has done it, but now, thanks to HPC, we see what it looks like. Hold on to your feet because NASA has released videos of wh Read more…
February 21, 2024
While 2023 was the year of GenAI, the adoption rates for GenAI did not match expectations. Most organizations are continuing to invest in GenAI but are yet to Read more…
May 8, 2024
Chip companies, once seen as engineering pure plays, are now at the center of geopolitical intrigue. Chip manufacturing firms, especially TSMC and Intel, have b Read more…
March 27, 2024
Pictures of Nvidia’s new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…
© 2024 HPCwire. All Rights Reserved. A Tabor Communications Publication
HPCwire is a registered trademark of Tabor Communications, Inc. Use of this site is governed by our Terms of Use and Privacy Policy.
Reproduction in whole or in part in any form or medium without express written permission of Tabor Communications, Inc. is prohibited.
