Maybe he's taking money from TSMC/Nvidia/Samsung/Intel.
Chinese semiconductor thread II
- Thread starter vincent
- Start date
Considering the article is behind a paywall, I think being able to at least access the actual contents would be a prerequisite before asking your question.
It will be nice is someone post the article.
I have my doubts that is what they said because digitimes is one those publications who writers like to put words in other people mouths, I haven't found any remark remotely similar to what digitimes says, neither in the CSIA page or in conference page or even in social media.
But let suppose that is true, so what is Huawei suppose to do? Wait for a miracle that change the hearts of US politicians? Huawei is suppose to let themselves die?
Another thing is again lets suppose is true, if these executives all compete in the same old nodes their companies are going to die, in order to survive they have to move up or specialize, they have no other option.
For example Naura and AMEC are increasingly competing with each other, in some cases overlapping tools, but it get even worst when you have new companies trying to grab some market share in the mature nodes, Naura and AMEC have no other option than to move up where the new companies can't compete that easily.
HH is having increasing competition in the mature node, so they have to move up to more advance nodes whatever they like it or not.
SMEE is having increasing competition in the packaging lithography market with a lot new companies making offerings in that sector, so they have to move up deep into IC manufacturing, they have no options.
That is the beauty of competition, companies who decide to get stagnant for profits they die as soon they competition catch up.
And get even worst, there zero guarantee that US export controls will not extend to mature nodes in the future. So if Chinese companies want to get stuck making mature chips then their best bet is to do it with 100% local equipment, materials and parts, that way they may have a hope they in the future they may be able to move up as Chinese toolmakers move up too.
broadsword
Brigadier
I don't have access to the article as well, I read a small snippet from the article someone posted on twitter, which seemed to be directly quoting the head. So happy to know more and frankly to be corrected. This statement seemed totally out of line with everything that one would expect.
Without the full article, I think the question shouldn't be "does anyone know why they made those remarks" and more "does anyone have the full article" to begin with.
While the media is still buzzing about Huawei joining hands with Semiconductor Manufacturing International (SMIC) to manufacture advanced process 7nm and even 5nm chips, on the same day that TSMC held its 2024 Technology Forum in Hsinchu, China's 26th IC Manufacturing Annual Conference and Supply Chain Innovation Forum was also held in Guangzhou during May 22-24.
Ye Tianchun, President of the Integrated Circuit Branch of the China Semiconductor Industry Association and Vice President and Secretary-General of the China Integrated Circuit Innovation Alliance delivered a keynote speech on "Path Innovation and Leapfrog Development" in the Chinese IC industry at the opening ceremony of the conference, highlighting the importance of continuous innovation to upgrade technologies at the mature nodes.
Ye analyzed the current state of development in China's IC industry across design, manufacturing, packaging and testing, equipment, materials, and components through various sets of industrial data. He pointed out five major blind spots in the Chinese industry, expressing concern that within 3 to 5 years, Chinese chips might regress to mid-to-low-end levels, worsening the situation.
He noted that the development approach of China's IC industry over the past decade has primarily been focused on "catching up with existing technological paths" and addressing "shortcomings." While this method allows for easy catch-up and fewer detours, it also risks falling into "path dependency," resulting in technological lag and a passive stance.
In this regard, Ye Tianchun believes that from a strategic perspective, "path dependency" is the biggest obstacle to the Chinese IC industry advancing to higher levels. Particularly on traditional technological paths, China's progress in advanced processes is being fully obstructed. If China cannot forge a new path, there is a risk of regressing to mid-to-low-end levels within 3 to 5 years.
Focusing on back-end and mature-node innovations
He emphasized that the external pressures will compel the Chinese IC industry to innovate its paths 7 to 8 years ahead of the global semiconductor industry.As Moore's Law approaches its physical limits at 2nm and 3nm, the global industry is considering innovations in product architecture. However, China needs to consider architectural innovation before reaching 7nm, which is why 3D system packaging has become so prominent in recent years.
Particularly, the difficulties encountered by China in front-end processes such as advanced wafer manufacturing will force China to innovate in back-end processes such as packaging and testing, creating opportunities for 3D heterogeneous integration system packaging, FDSOI, and FinFET technologies.
However, when discussing the "path innovation strategy," industry insiders in China often fall into common cognitive blind spots. Ye Tianchun pointed out that one such blind spot is China's 3D system packaging being detached from product design architecture innovation, which is like trying to innovate behind closed doors.
Chiplets have become very popular in the semiconductor industry, but their essence lies in system packaging. To achieve this, the focus should not be on the processing side but on the design side, requiring cooperation between IC designers and application system manufacturers with semiconductor packaging and testing companies to form solutions based on product architecture innovation, said Ye.
Ye Tianchun directly pointed out another common blind spot inside and outside the Chinese industry, which is focusing solely on so-called "advanced processes" while ignoring "mature process" innovative features. He emphasized that developing special processes within mature processes to advance from mid-to-low end to high-end is a critical task for path innovation.
Since the equipment and technologies for producing advanced nodes are under export control by the United States, China's opportunities are in the mature and legacy nodes. With dozens of new fabs being built in China to produce mature chips to increase output significantly in the next few years, China needs to upgrade the quality of its chips to gain market dominance. Otherwise, the oversupply of low-quality mature chips will result in a vicious cycle of cut-throat price competition among local producers.
broadsword
Brigadier
Exactly. He is not telling the industry to forget about the advanced nodes.
Thanks to @Awenumick and @Topazchen.