On May 15, 2020, the U.S. Department of Commerce issued an announcement that Huawei was "undermining" the list of entities and therefore was restricted from using U.S. technology software to design and produce semiconductors.
The news has raised concerns about Huawei and the Chinese semiconductor industry.
So, in the face of the U.S. technology hegemony, how Huawei should face, how China's semiconductor industry should move towards independent and controllable, these issues have also caused widespread concern.
As a result, Xinda Securities electronics industry chief analyst Fang Jing and Huawei strategic advisor Mr. Meng conducted an industry analysis report on this incident.
Good afternoon, investors!
Today's conference call features Mr. Meng, Huawei's strategic advisor and former Blue Army commander, who joins us to interpret the May 15 announcement from the U.S. Department of Commerce.
The Blue Army is a Huawei-specific organization that is directly under the Strategic Marketing Department of the core functional platform. The main role is to provide strategic advice to Ren Zhengfei and Huawei EMT (Executive Management Team) by simulating and researching competitors.
I. Interpretation of the U.S. Department of Commerce Announcement
Let's start with the basic situation, on May 15, the U.S. Department of Commerce issued two consecutive announcements, the first is the last extension of the temporary license for 90 days until August 14, this extension is different from the market interpretation and actually refers to the extension of Huawei can continue to provide services to U.S. customers.
Since Huawei doesn't have much business in the U.S., this temporary license extension, though it's the last one, will have little impact.
But a second announcement followed, escalating chip controls on Huawei, which the U.S. Department of Commerce said "undermines" the list of entities and therefore wants to restrict Huawei from using U.S. technology software to design and produce semiconductors. (Restricting Huawei's ability to use U.S. technology and software to design and manufacture its semiconductors abroad).
How exactly is it limited? The United States has expanded its controls to make the following materials subject to EAR (export control regulations).
(1) Huawei or its affiliates on the entity list (e.g. Hays) use products designed using software and technology from the Control Control List (CCL), such as semiconductors. (Restrictions mainly for EDA)
(2) Products such as chips designed by Huawei or its affiliates on the entity list (e.g., Hays) that are manufactured outside the U.S. and use equipment on the Control Control List (CCL). A permit is required at the time of import or export. (Restrictions mainly for fabs)
At the same time, in order to avoid the impact on semiconductor equipment companies and fabs, the U.S. Department of Commerce said that any production step for items based on Huawei design specifications as of May 15, 2020, which has been initiated on May 15, 2020, shipments to Huawei will not be affected for 120 days after this regulation takes effect. (This also means that impromptu orders are not allowed, and that the already cast film is not affected)
ii. impact analysis
The market is extremely concerned about this restriction plan for Huawei, but we think the impact on the software is relatively limited, because the US EDA company stopped working with Huawei last year and cut off the upgrade, now Hys is using the old version of EDA for product design, without restrictions.
At the same time, Huawei's chip design cooperation with ST can also be outsourced to a certain extent to solve the EDA problem.
And for fabs, it may be a bigger challenge. Of course, not too pessimistic. First of all, after the regulations came into effect on May 15, there is a 120-day buffer period, buffer period TSMC and other manufacturers can still ship to Huawei; secondly, we have also stressed that Huawei plus single and other matters, the company still has a rainy day plan.
So from an optimistic point of view, the U.S. restriction plan on the one hand, left more space for mediation, related semiconductor manufacturers and potential trade disputes affected by the terminal manufacturers will vigorously lobby the U.S. government within the deadline, the leaders of both countries also have room for urgent consultations. At the same time, the U.S. restrictions on fabs are more of a requirement to apply for a license, rather than a one-size-fits-all, and will probably continue to extend fab supply licenses within a certain time limit.
Huawei's global base station market share reached 30.9% in 2018, and it has many layouts in the European market.
On May 5, Deutsche Telekom publicly backed Huawei, saying that if Germany wanted to achieve its goal of building 2,000 new LTE base stations by the end of this year, it needed Huawei's participation, and if it rejected Huawei, Deutsche Telekom would not be able to solve the 5G signal coverage problem quickly. If the U.S. blatantly cuts off Huawei's supply chain across the board, it will bring irreparable losses to European operators' 5G deployments, which will also bring a global backlash!
iii. guest speakers
The U.S. Commerce Department's announcement of restrictions on Huawei's use of U.S. technology software to design and manufacture semiconductors is also a plan, not a resolution. And it goes on to say something like this: "For 120 days after the effective date of this regulation, shipments from fabs to Huawei will not be affected." Is it possible that this plan will become a formal regulation? I think it's possible. But it's still possible to continue to extend the period when it gets near 120 days.
For countries like Germany and the UK, the US can always say no to Huawei stuff, for third countries, it's intolerable. In fact, the United States is still very cautious, this deterrence is more than practical, good for Trump's vote, in addition, will also increase the concern of Huawei's customers, may be on Huawei's order chop. This thing is actually the U.S. tearing up its business credibility, the signed terms are all torn up, and the U.S. wants to do this thing to take its business credibility.
It is important to analyze the U.S. sanctions against Huawei, or the U.S. decision-making model.
1) In fact the US President is very powerful externally, the nature of the President is very important to the game between the two countries, Trump's character this person is able to sacrifice any interest for personal gain, including the interests of the US. For example, if we are supposed to fight against the epidemic, we should work with the big countries to get some protective clothing and testing reagents, and we only need to consider deaths within 200 000 as KPI.
2) Trump does things arbitrarily, sanctioning Huawei is not the plan from the beginning. Huawei this matter, first hit ZTE, then prepared for Huawei for a year, then arrested his daughter, not in line with the normal combat methods, and later on, including the extension to Huawei, has signed four times 90-day extension period, during which Huawei has replaced the parts. Sanctioning Huawei not only damages the credibility of the United States, but also hurts the interests of the industry chain.
3) Trump is a self-serving man, it's a tool to get public opinion, to get votes. From this perspective one can understand when to make a move against Huawei, if a move against the whole of China would be to hit on the cotton, hitting Huawei is more figurative, every hit Huawei can increase his approval rating.
Analyzing Trump's character, the set-up of hitting Huawei was wrong, by hitting Huawei to get votes. This is an election year, and if Trump's approval ratings fall behind Biden before the election, there could be extreme measures to gain votes.
It's safe to say that no matter what the US does to Huawei, Huawei will survive, and it's taken 1 year to prove it. The United States does not have unlimited firing power and the two sides may still talk. It's possible to eventually turn the motion into a resolution, but there are 120 days for Huawei to prepare, and the US has defaulted to the fact that it can't beat Huawei to death.
So this thing, in the grand scheme of things, is Trump's strategy to get votes.
iv. report q&a
Q: How far does the ban affect Huawei's 5G base station, the limit assumes no way to get the 5nm base station chip from TSMC, how does Huawei simulate the worst possible state, what is the impact on 5G base station performance?
A (Moderator): The conference call just now also said that Hays has urgently added a 5nm order to TSMC, and if it is completely banned in the short term, it will actually maintain the supply of base stations for at least 3 to 6 months.
We can also see that on May 6, the United States signed new regulations that allow American companies and Huawei to customize the 5G standard.
It is contradictory from the underlying logic that a US company can work with Huawei and issue the Commerce Department, and now it is also the Commerce Department that issues the upgrades to Huawei's restrictions. We still think the United States wants more to put this thing on the table.
Q: What about other companies in the Huawei industry chain?
A (Moderator): People are worried that Huawei's industrial chain will be impacted, first of all emphasize the 120-day buffer period, Huawei can adjust the design, we think will increase the proportion of domestic chip use, while finding a spare tire for Hays. For example, heath originally did a lot of power chips, mobile phone processor PMU, originally heath himself do, the next will increase the proportion of such enterprises as the Holy State, to help it do some more meaningful alternative chips. There will certainly be a shock in the short term, when stocks are depleted, the back-to-back transition period will be slightly more difficult, but then there will certainly be a boost to domestic substitution.
A (Guest): Overall, there is not much impact on the Huawei industry chain, Huawei will not die, only possibly the next period of time to put the phone to do backward points. As long as Huawei is not dead, the supply chain will supply what it should supply, growing together more closely and in synergy. So the effects are emotional, not substantive.
Q: Can the U.S. have long arm jurisdiction over Japanese semiconductor supplies?
A (Guest): It can't be done, the United States has the power, and the power is not to do whatever you want.
Q: What happens if the 120 days are not extended?
A (Guest): The U.S. would be more cautious because of the goodwill involved, but there is that motivation, depending on Trump's approval rating.
Q: Is it extremely unlikely that other countries will follow the United States entirely?
A (Guest): Your business partner can tear up the contract at any time, meaning that buying his stuff and invalidating the business contract is a violation of basic credibility.
Q: What leverage does China have to threaten the United States?
A (Guest): We don't have complete heights in the entire chain, what we do have is a vast market, but it's not enough to threaten the United States, it's defensible. The market is big enough, and a lot of the dirty work in the chain is being done domestically, and it would be detrimental to the U.S. to have a full-scale confrontation, so eventually it will be negotiated.
Q: Will Huawei's business shrink significantly in this form?
A (Guest): It won't die, and it won't shrink too small. Huawei's industry position in many areas is still solid, and it's hard to tell how far it will shrink next. But I agree with Cinda that it's important to find a backup for Heath.
Q: How much pressure can the U.S. exert on ally countries? How much influence?
A (Guest): Every country is different, every country's diplomacy depends on its strength, its leaders, and the United States cannot turn the will of the state into the will of the corporation, which is a problem for the entire United States national system. If Huawei collapses, TSMC's big customers will be gone, which is something who doesn't want to see.
Q: The market has a kind of view on how Huawei will respond to the U.S. sanctions, if Hays breaks up Huawei can stage to alleviate this contradiction inherent logic what? Where is the point of Huawei's reluctance to split if the contradiction is properly delayed?
A (Guest): the Trump team hit Huawei all for approval ratings and votes, it's too far away to talk about a split.
A (Moderator): The possibility of the whole Haith divestiture is very small, but we can look at the recent changes at Huawei, there have been recent news reports of corporate changes at Huawei's subsidiaries, Huawei also wants to make internal adjustments, and they will be prepared for that. If the United States is serious, it can have a delaying and deferring effect and eventually return to the negotiating table. The most pessimistic scenario is that Huawei Hys will eventually become an IP company, retaining the core IP licensing business, and the R&D staff will be dismantled to support other domestic chip companies.
Q: The gap between SMIC's 14nm process and TSMC, and the 2% revenue share of high-end support in 2020Q1, how much room is there for improvement?
A (Moderator): SMIC is also promoting N+1 in addition to 14nm, and as I mentioned in my weekly industry tracking update, in fact, in the first quarter, SMIC's 14nm is an RF Transceiver, and the volume is relatively small, so capacity utilization is not high. Recently the Kirin 710A has also started batching, which will pull the crop movement rate in the future. However the 710 is a 2018 launch processor with a limited lifecycle total that can more than look forward to the upgraded version of the 7 series processor, as well as the Low Cost version of the Kirin 820 that is currently in test validation.
However, 14nm revenue accounted for more than 10% by 2021, domestic replacement is a gradual process, but still have to respect the laws of the industry. the difference between SMIC and international leaders can not be underestimated.
Q: SMIC has 14nm capacity, but it's not big. If it supplies Huawei, will the US ban SMIC?
A (Moderator): We can't judge whether it will be banned, we think that now is more the time to put together the industry chain speed, SMIC has indicated in the preliminary legal meeting that it will increase the capex by $1.1 billion this year, purchase more equipment, more capacity is always right. Moreover, Huawei has a massive stock-up order in advance, which according to the most pessimistic expectations, is also enough to support March to June.
In response to this report, Leiphone.com also interviewed the chief analyst Fang Jing, the interview mainly related to the Huawei in the case of the U.S. suppression of Huawei's response and survival strategy, and China's semiconductor industry independent and controlled future development, which can have a more three-dimensional understanding of the suppression of Huawei.
The following is a transcript of this interview.
I. What is the real purpose of the US chip regulation of Huawei?
The U.S. restrictions on Huawei are explicitly Huawei's threat to national security, but are actually a repression of China's technology industry.
If one looks back at the US trade war against Japan in the 1970s, one finds that it was a consistent US ploy, starting with calling on the American public to smash Panasonic's radios, then trapping Panasonic executives, then filing endless lawsuits and injunctions against Panasonic, and finally forcing it to abandon its semiconductor business.
If you compare what the United States is doing to Huawei today, you will find that it is a repeat of yesterday, and the methods are the same.
What is the impact of this ban on Huawei's use of EDA?
With EDA, the impact on Huawei is not too great, because Huawei can continue to use the old version of EDA, but not the new version. Of course, the autonomy of the EDA is also very important.
In fact, a number of high-quality EDA companies have emerged in China, some of which have achieved a follow-on to the United States, or even reached the level of catching up, for example, Huada Jiutian, which has done very characteristic at the simulation level, and many listed companies have cooperation with them.
3, TSMC's exports are affected, how should Huawei respond?
Huawei will import more domestic chip design companies and strengthen cooperation with the eco-chain. Of course, it is also possible for HSI to turn the whole thing into something that can empower domestic chip companies. It's also because there's a big gap in our manufacturing, and if a company as strong as TSMC, it wouldn't be so passive.
Of course, personally, I don't think this thing is actually that pessimistic. As can be seen, TSMC is not completely banned, it can internally review the application for permission. In addition to this, Huawei itself has made a lot of efforts, such as actively stocking up on core devices, with at least half a year's worth of inventory.
4. Is the U.S. ban related to TSMC's plant announcement?
Previously, the market thought it was related, seeing it as a gesture to the US to keep Huawei. But I now think it was just a coincidence in time.
What TSMC announced was to build a 5nm plant in the US, and although it looks advanced now, the plant is to be built in 2024, and 5nm won't be advanced by then. And with only 20,000 slices of monthly capacity, this plant is so small that its significance is more on a symbolic level, more like a Trump administration performance project.
Fifth, the current development of semiconductor technology is difficult to leave the United States of America's technology, if the United States under a deadly hand, what will eventually lead to a result?
I don't think this pessimistic scenario will play out. The same was true of Panasonic when it was first introduced, and although it took a lot of beating, it eventually held up. Huawei will certainly survive, and the most serious scenario is for Hays to become an IP licensing company, spun off, playing a role of burning itself and lighting up the industry.
The US is currently targeting one Huawei company because Huawei threatens US technology hegemony; once we are strong enough in the industry chain, say we have ten companies as strong as Huawei, we don't have to care about US attitudes.
VI. China's semiconductor industry, how is the process of independent localization?
The autonomous and controllable Chinese electronics industry is cognitive in stages. Our earliest was brand recognition, such as Lenovo, Hisense and so on, and then realized that semiconductor design is very important, so we have such companies as Hisense, Huideng, and so on, and then also not enough, we need to have a wafer fab, so we have such high-quality listed companies as SMIC, Hua Hong. But it now turns out that upstream equipment and materials are also important, and as our Neutrino company, North Huachron is catching up.
So, autonomous and controllable is one ring after another, and the further upstream you go, the more control you have. We actually have a strong enough discourse downstream, but the downstream is still constrained by the upstream, so we have gaps at the upstream level.
7. What progress has China made in independent research and development of photolithography machines?
At the moment, Shanghai Microelectronics is something to look forward to. But the high point of the lithography machine was in the Netherlands, not completely controlled by the United States.
Lithographs are important, but what we lack is an all-around strong upstream industry. For example, the United States has applied materials, PanLin Semiconductor, comprehensive layout PVD, CVD, etching machine, cleaning machine and a range of other equipment, if we can also be strong in these devices, there is no need to worry.
Of course, when we're strong enough, we can license each other; while the lithographs are also using American technology, the U.S. is also applying technology from other countries. So, if one day we are strong enough that ASML is applied to our technology, there will be a bottom.
For example, in the earliest days, our panelists still had to pay royalties to manufacturers such as Samsung, but now they no longer do. Why? Because of the panel we have achieved follow through and in some areas achieved beyond.
So, it's not that the Netherlands has to listen to the US, it's that we're not strong enough to let the lithography machine use our patents.
Eight, what is the impact of the ban on the Chinese semiconductor industry?
There is no doubt that it will push China's semiconductor industry backwards, especially in terms of upstream equipment and materials.
In fact, China has invested a lot of money in the semiconductor field over the years, but we still frequently see related companies being suppressed by the United States. Because semiconductor is not a single industry chain, this chain is very long, we are gradually moving up from the application side, it takes time, we have made a lot of progress, just not enough.
IX. What do you think about reports that China may take some countermeasures involving Apple, Qualcomm and Boeing?
I think Boeing Cisco is possible, but not Apple Qualcomm.
Companies like Apple are actually helpful to China, it has driven domestic employment, helping China to cultivate a large number of technical personnel and manufacturing practitioners, many listed companies are the industrial chain behind Apple, so it also involves thousands of jobs. The same is true of Qualcomm, which is used by many domestic mobile phone companies, such as Xiaomi, OPPO, vivo and others.
We do not need to be so pessimistic about the electronic sector, especially in relation to the national livelihood of the consumer electronics industry, is to move forward; not that the United States sanctions, will create absolute confrontation. I think there's still hope for reconciliation through negotiation, not confrontation.
In doing so, the U.S. may also be a vote consideration; there is really no need to be so pessimistic in light of the Japanese experience.
X. Is it possible to form two technology industry systems in China and the United States in the future?
It's possible.
But American technology comes first, and we may still follow American technology at first, but later we may also develop our own technical characteristics.