Back in 2019, I wrote a two–part article on China’s electronic design automation (EDA) industry. I argued that despite some promising domestic companies, China would not be independent of the dominant US- and Europe-produced tools anytime soon. It is now six years later, and this is still the case. Back then over 90% of EDA tool sales in China came from Synopsys, Cadence, and Siemens (the big three). Over the past five to six years this has reduced to around 80%, still 10% higher than their global market share of around 70%. Although market share is not the only measure, it does provide an idea as to how much Chinese fabless companies rely on imported design tools.
Since 2020, certain entity list Chinese companies like Huawei have been restricted from licensing EDA tools from US suppliers, but in late May this year, the US announced that these restrictions would be extended to all Chinese companies. New tool sales have reportedly been stopped, and Chinese customers now have limited access to support portals.
Why this matters
Right now, Chinese users can continue using the tools they have already licensed but maintenance and updates will cease once contracts end, meaning customers would have to find ways to maintain and update them themselves. Chinese fabless companies whose contracts end soon could face design delays, whereas others may have a longer time to plan for alternatives.
China actually has more EDA tools companies than the rest of the world combined, with at least 60 such companies. Domestic companies cover nearly all tool categories. So, in theory, out of any other country in the world it is best placed to deal with such restrictions. However, among the tools already developed by domestic companies, those that have passed customer demos are mainly focused on analog or mature nodes, not advanced nodes.
What’s more, even though Huada Empyrean and Primarius Technologies may claim to support 7nm or even 5nm platforms, they only cover part of the process. Having domestic EDA tools is one thing, but whether they are usable and easy to use is another.
Often the biggest challenge with EDA is that it’s never developed or optimized purely in labs. It is refined in the real world. Without use in the industry it is hard for a tool to improve. Even if a tool is painstakingly improved, it is bound to be riddled with bugs when customers first demo it. Real customers need to use it and hit these bugs for improvements to happen.
Opportunity
In theory then, these restrictions are an opportunity for China’s EDA companies. They now have the chance to gain more real-world experience. If domestic fabless companies have no other option but to adopt domestic tools, then the more people will use them, and the better they’ll become. These restrictions could therefore be a blessing for China’s EDA companies, while simultaneously being a nightmare for China’s fabless companies. Indeed Empyrean now aims to ascend to the top tier of EDA providers, capitalising on the struggles faced by its US competitors in selling to China because of new export controls on chip-design software, according to Yu Han a senior market director at Empyrean, who revealed this goal at the World Semiconductor Conference in Nanjing last week.
Not only is this a market opportunity for domestic EDA companies, but also an investment opportunity. Investors are sure to see the unreliability of US suppliers as an opportunity for domestic companies to gain market share and boost revenues.
Challenges
As I said though, this could become a nightmare for fabless companies. While some out there, especially in China, might argue that China can do everything itself, this just isn’t the case. In core areas such as EDA, materials, and equipment, Chinese companies, despite advances, still rely heavily on imported technology.
The truth is that currently even the most celebrated domestic chips such as Hisilicon’s Kirin and Ascend, and Cambricon’s Siyuan, all rely on imported EDA tools, equipment, and materials. If all imports were actually cut off, China would not be able to produce any advanced chip, even mature chips would be a struggle. China needs to understand this. Over confidence could put it in a difficult position. I believe this is well understood though. While Chinese media may claim sanctions are good for China’s industry, and in some cases like EDA they could be positive for that specific sector, overall, they are not. Why else would Chinese companies continue to use shell companies and other methods to import that technology they require?
Conclusions
EDA tool restrictions, if enforced, will be bad for the big three, bad for the Chinese semiconductor industry, but good for China’s domestic EDA tool suppliers.
Chinese firms like Huada Empyrean and Primarius are working toward full-stack EDA platforms. However, for the time being, no imported EDA tools, would be a catastrophe for Chinese chips.
China’s semiconductor industry can still operate today despite US pressure by expending massive resources and paying extremely high costs to find workarounds. What worries the Chinese industry is that it could be cut off from all foreign technology before it has been able to catch up. It hopes to be able to continue to import foreign semiconductor technology like EDA tools, while gradually building up its own capabilities. Any sudden cut-off from foreign technology before it is ready would be a huge issue.
China found workarounds when it was blocked from advanced equipment, and one would be ill advised to suggest they could not do the same with EDA given the advances made. However, workarounds are workarounds – they will never be leading edge, or lowest cost, and a lot of waste is involved in keeping oneself from drowning. China needs to somehow innovate and advance while maintaining access to foreign semiconductor technologies, something it may manage to do with the various levers it has at its disposal in wider trade negotiations.
