By Jeff Ferry, CPA Chief Economist
Despite the COVID-19 crisis, Chinese technology company Huawei is having a good spring.
In March, it reported that 2019 revenue rose 19 percent to $123 billion, a new high. In the first quarter of this year, Huawei’s chip division HiSilicon surpassed Qualcomm to become number one in the China market in sales of chips for smartphones. Meanwhile, in 5G wireless network systems, Huawei continues to be the undisputed leader, having announced in April that it won a 57 percent share of a mammoth $5.2 billion contract with China Mobile to build 232,000 base stations across China.
Nor are Huawei’s successes limited to telecom equipment. It is number one in smartphone sales in China with a 39 percent share in Q1 of this year and number two worldwide, having surpassed Apple last year to take a 17.6 percent share of the global market. In online services, it recently launched a music streaming service with features like surround sound and “Party Mode,” likely to be popular with the same young Europeans who buy its Mate brand smartphones. Outside of telecom, Huawei is number one in selling solar inverters, an $8 billion market, despite being barred from the US market.
Sun Tzu said: “Know your enemy and you will win a hundred battles.” Huawei is likely the most successful Chinese company in the world. While its early days were marked by widespread intellectual property theft and cheap underperforming products, today Huawei is a world-class competitor, leading many markets and spreading its wings into new ones despite intense opposition from the Trump administration and some other national governments. It’s true that thirty years of government subsidies and compliant employees give it an advantage western companies lack, but we must not underestimate our competitor.
Twenty years ago, when I was young (well, younger than I am today) and working for Nortel Networks, Huawei was a small ($3 billion) upstart making inferior, copycat products. Cisco was the leading networking company with seven times the revenue of Huawei. Nortel had $20 billion in revenue, provided a large chunk of the gear for the AT&T and Verizon networks, and was the world leader in fiber-optic networks. Today, Nortel is bankrupt and gone, while Huawei is world leader in fiber-optic networks and wireless networks—and twice the size of Cisco.
Huawei is a huge threat to the US economy and US national security. Even though its equipment is rarely deployed in the US, its wide use outside the US in the networks and by consumers in most of the markets of our major allies is a threat. You only have to read the tale of the African Union’s in-house network to see that Chinese networks are a security threat. Equally important, Huawei is a threat to our chip industry, which is the heart of US technological leadership. As Huawei and China move from becoming the leading purchasers of chips to leading developers and manufacturers of their own chips made on Chinese soil, it threatens our technological leadership, our defense capabilities, and the economic health of leading US chip companies like Intel, Qualcomm, and Broadcom, already too dependent on the Chinese market. With annual revenues at $10 billion a year and now selling chips to other equipment makers, Huawei’s HiSilicon is now the world’s 10th largest chipmaker.
In that context, whether or not the Trump administration’s embargo of chip sales to Huawei is effective is a secondary issue: in five to ten years, Huawei (and other Chinese equipment makers) won’t need US chips. They will follow the dictates of Made in China 2025 and buy only Chinese-produced chips. Unless US chipmakers have US and other non-Chinese customers to turn to, the US industry will struggle.
A new bill, the Utilizing Strategic Allied (USA) Telecommunications Act of 2020, sponsored by a bipartisan group of congress members led by Rep. Frank Pallone (D-NJ) and Rep. Greg Walden (R-OR), is to be welcomed. It allocates $750 million for grants to network equipment vendors to support 5G wireless networks using OpenRAN technology. It is similar to the Burr-Warner bill in the Senate (S. 3189).
But the bill is too narrow, in two respects. First it is limited to one technology, OpenRAN, which may or may not turn out to be a winner. Secondly, it talks about beefing up the supply chain for 5G wireless networks, but misses the point that the key missing ingredient is a new network vendor, and preferably a US-based network vendor.
Take the second point first. The reason why foreign telecom companies and governments have largely rejected Trump administration pleas to ban Huawei from their networks is that such a move would restrict their choices for 5G networks to just two, Ericsson and Nokia. Two choices are insufficient to ensure aggressive competition, rapid technological progress, and good customer service. In the words of Orange Belgium CEO Michael Trabbia: “If we limit, with the US trying to ban Chinese vendors, we will end up with only two vendors, and that is an issue.” You can also watch Boris Johnson, speaking for the UK telecom industry, saying the same thing here.
I spoke to an old friend of mine who is a senior engineer on the team that evaluates networking systems for a major European telco. Like many European technologists, he would prefer not to do business with China, because of the way they exploit European markets, while keeping their own market largely closed. But his employer uses Huawei. “Huawei performs very well, both from a technological standpoint and contractually.” His employer is keeping its dependence on Huawei from rising too high, but it does not want to be too dependent on Ericsson and Nokia either. “What is needed are more vendors offering quality solutions,” he says.
That’s the top priority: more vendors who can compete head-to-head with Huawei, Ericsson and Nokia.
What about OpenRAN? OpenRAN is a technology that will enable reduced costs and more flexibility in the wireless network, which is good. But from the point of view of US strategic interests, its value is that it could be a stepping stone to allow new 5G network vendors to emerge. At present there are three US-based startups that are candidates to become complete 5G network vendors: Parallel Wireless, Altiostar, and Mavenir. OpenRAN will help them accelerate their development, which is good. But it’s not enough.
OpenRAN is driven largely by Microsoft and VMware, a division of Dell. Those two companies want to shoehorn themselves into the $200 billion a year telecom equipment market, by getting telecom networks to use “cloud” i.e. server virtualization technology. There’s nothing wrong with that; it would reduce the cost of the network. But it is not a complete solution, so in itself OpenRAN doesn’t offer a new alternative to the existing vendors. At my friend’s telco in Europe, they are actively testing OpenRAN systems. He tells me: “OpenRAN is interesting. But the ecosystem is not large enough yet. And they seem to want us, the telcos, to put the pieces together ourselves. And that is a problem.” By problem he means unlikely to happen unless and until a single vendor can offer it as part of a complete 5G system.
Technology history tells us that while all technology companies and their customers praise “openness”, technological breakthroughs usually happen not from openness, but from great products that are faster, better, and cheaper than the competition. Example: Cisco has been criticized for 30 years for selling routers based on closed software systems. Yet it’s still the world leader in Internet routers. Customers love its “closed” routers. However, it does have one new, successful competitor today: Arista Networks. Arista’s products are not open either, but they are blazingly fast. Why? Because they were developed by a brilliant serial entrepreneur who in my humble opinion is the world’s outstanding hardware engineer, bar none. Silicon Valley history is that we have led the world in technological progress because at any given moment there are 10 startups trying to build something that is faster, better cheaper. Nine will fail. One will succeed, and “change the world” (or at least change the technology industry).
According to Earl Lum, a former wireless engineer who is today a consultant, the Huawei 5G radio (the box that goes on the cell tower) is, at 25 kg, some 37 percent lighter than the 5G radio from either Nokia or Ericsson. Weight is a good proxy for capital cost, operating cost, and power consumption. All of these systems are too expensive and power-consumptive compared to what they will be five years from today. But the non-Huawei ones are the worst. “If we did deploy 5G nationwide in the US today, we would need a new nuclear power station to power them all,” Lum tells me.
What is needed are technological breakthroughs in design that take cost out of the communication chain, not just virtualization of the software. Parallel Wireless is working on just such breakthroughs. (Full disclosure: Parallel Wireless is a member of our Coalition.) But we need more companies working on more such challenges. And pursuing diverse approaches. Because many will fail but a few will win. And pardon me if I’m skeptical, but any approach run by a committee dominated by Microsoft and Dell is unlikely to develop the breakthroughs we need. It can help, but it won’t solve the problem.
My other concern about Pallone-Walden is that it sets up a committee to advise on disbursement of the $750 million, with members from the FCC, the Department of Commerce’s NTIA agency, and others. I worry we are setting ourselves up to fail. We will distribute some money, we will build a “test bed” where people can test their equipment, but we won’t have enough money or brains focused on the real challenge. In the end, we won’t have a challenger to Huawei, but everybody on the committee and the tech participants will be able to say they tried their best.
Washington works by committee. Everybody meets, talks, compromises, and talks some more. Silicon Valley is the opposite. On a business level, Silicon Valley excels at one thing: uniting an entire company around a single objective. At a young startup, everybody is focused on building a product that works before the cash runs out. (I worked at two, one a huge failure and the other a huge success.) At a well-run startup, nothing else matters. We need that Silicon Valley startup mentality in 5G wireless today.
In my travels in Washington I find very few people who understand the huge challenge of building a product that can compete with a real-world-hardened product from a $123 billion gorilla. Instead of Washington or tech bureaucrats, what we need is to appoint a committee of industry veterans, give them a set period, say five years, and give them marching orders and the one goal that matters: you will disburse funding to create a successful (i.e. profitable) new US-based competitor to Huawei, Ericsson, and Nokia. Or two new competitors.
They should be men and women with proven track records, perhaps retired. Eric Schmidt former chairman of Google, is a good example of the sort of person we need. Eric has captured Washington’s attention with his interest in advising the Pentagon on artificial intelligence software and with this article: “I Used to Run Google. Silicon Valley Could Lose to China.”
I can think of half a dozen Silicon Valley veterans like Eric who could join this committee. Some, like him, are Democrats, and other are Republicans. It doesn’t matter. We would not be hiring them for their opinions. We would be hiring them because they answered “yes” to this question: “The USA wants to give you $750 million and we want you to use it to create one or two competitors to Huawei because we need a complete wireless networking technology supply chain in the US. Can you do it?”
Of course they should have no financial interest in the outcome. They will be doing this because they are already wealthy and they want to do something for their country. Like the industrialists FDR hired to build America’s military production in WWII, they will be “dollar a year” men and women. The process should be completely transparent. They will be able to call on the support of the Pentagon, of In-Q-Tel, of the national security community, of the FCC, and of AT&T and other large telcos who will all agree to test the new products, and, if they are good enough, buy them.
Here in summary is my proposal for a new US 5G wireless telecom strategy:
- Recognize the true scale of the problem and the challenge.
- Recognize that the only way to build a non-China worldwide networking ecosystem is to create new US-based competitors to the Chinese vendors (while keeping the existing competitors in business).
- There are startups and potential startups (and perhaps some mid-sized companies in adjacent businesses) that could grow into these new competitors.
- The federal government must take the lead with an expert committee that can lead and manage the US fightback on wireless networks.
- The expert committee must be given a singleminded focus, and time, money, and support.
Then step back, and watch the US tech industry work its magic, as it has so many times in the past.
Jeff Ferry is Chief Economist at the Coalition for a Prosperous America. Before joining CPA, he spent 18 years in the networking industry. Follow him at @menloferry.