Starved of chips, Huawei pivots to cars and software

Few people today would be aware of Nokia's 19th-century origins as the operator of a pulp mill in Tampere. The evolution of the Finnish firm into one of the world's biggest makers of telecom equipment has taken more than 150 years. Transformation could make one of China's largest pandas similarly unrecognizable in a much shorter period. Battered by US sanctions, Huawei is shedding its identity as a telecom hardware business.

The Chinese equipment giant aims to reinvent itself as a software company serving industries besides the telecom sector. It already boasts the number-two spot in the Chinese market for public cloud services and has established its young enterprise division as the company's growth engine. Enterprise sales last year were up 23%, to more than 100 billion Chinese yuan (US$15.3 billion), and account for more than 11% of the total. China's automotive sector is a prime target.

The emphasis on software is not surprising. US sanctions introduced under Donald Trump and maintained by Joe Biden have succeeded in cutting Huawei off from important hardware vendors such as TSMC, a Taiwanese foundry that makes the chips designed by Huawei-owned HiSilicon. Software is less susceptible to a US attack.


Huawei's Eric Xu (right) pictured next to his translator in Shenzhen today.

Huawei's hope is twofold. By improving its software capabilities, it aims to be less dependent on chip technology for competitive products, said Eric Xu, one of its rotating bosses, at an analyst summit today. The implication is that it would be able to rely on less sophisticated chips without sacrificing performance. For a state-of-the-art transistor measuring just seven nanometers, Huawei's only options today are TSMC and South Korea's Samsung, both of which are subject to US sanctions. Chips incorporating larger transistors are more widely available.

At the same time, Huawei has promised additional investment in businesses that are less reliant on these advanced process techniques used in chipmaking. Gadgets provided to numerous industries as well as some communications chips do not require the deep sub-micron technology that goes into high-end smartphones and 5G equipment.

"My understanding is that there are lots of chips you can make at 180 nanometers that would be perfectly functional for a communications network," says Andy Sellars, the strategic development director at the Compound Semiconductor Applications Catapult, a not-for-profit research and technology association that advises UK industry and government on semiconductor strategy.

Since the end of 2018, Huawei claims to have pumped about $2 billion dollars into boosting its software capabilities. "Overall software capabilities have seen remarkable progress and we will continue," said Xu today. The goal, he told analysts and reporters, is to boost the share of revenues that Huawei collects from software and services.

More like IBM than Ericsson

The strategic pivot could ultimately make Huawei look far more like a Chinese IBM or Microsoft than a rival to Nokia and Ericsson, says John Strand, the CEO of Danish advisory group Strand Consult. "They are saying if we cannot sell telecom infrastructure we will go after IT solutions because there are more limited possibilities of putting regulation out there," he tells Light Reading. "The question is whether a serious pharmaceuticals company or bank would put data into Chinese company solutions, but it doesn't matter if you look at the whole market of China."

The automotive sector alone neatly encapsulates the shift Huawei is now attempting. Its intelligent automotive solutions (IAS) business, a part of the enterprise group, is already as fully formed as the better-known devices unit, said Xu, with its own sales, marketing and delivery capabilities. This year, Huawei plans to spend more than $1 billion on research and development at IAS, a figure that would equal almost a twentieth of Huawei's entire R&D budget last year.

Want to know more about 5G? Check out our dedicated 5G content channel here on Light Reading.

"The Chinese market is a market of 30 million vehicles per year and that number will grow in future," said Xu. "Even if we do not work on markets outside China, that is a big-enough business for Huawei." In a further sign of hardware disinterest, he scotched talk of Huawei producing its own cars. And as if to emphasize that automotive success is not contingent on what happens to Huawei's telecom business, he even played down the importance of 5G. "If a vehicle can only realize autonomous driving through 5G, there are problems," he said. "That would raise a very high bar for mobile service providers."

The plan is to form partnerships with carmakers on newly branded vehicles that will incorporate Huawei's software. Under a deal with Beijing Automotive, a "whole series of products" are promised starting in the fourth quarter of this year. Another agreement with Guangzhou Automobile envisages new vehicles by 2024. Huawei says it is also in talks with German and Japanese carmakers, although Xu told analysts that "we won't have many partners for this deep collaboration model."

Separately, work on its Harmony operating system is the best indicator of Huawei's efforts to build software platforms that could support entire industries. Thanks to Chinese interest, Harmony already ranks as the world's third-biggest device operating system, behind Android and Apple's iOS, with 2.3 million registered developers, including 300,000 outside China. Even if it ultimately flops elsewhere, Chinese adoption could be sufficient to establish it as a major force.

Geopolitical rifts

None of this is likely to worry Microsoft or IBM, neither of which does a huge amount of business in China. The real losers could be the Chinese BAT trio of Baidu, Alibaba and Tencent, says Strand, as Huawei lurches from telecom into their territory. The rapid growth of its enterprise business last year within China would clearly have left Chinese IT rivals with a smaller opportunity.

Huawei's pivot is further evidence that a previously interconnected and global telecom and technology industry is starting to break apart. Until now, countries have had their own specialities, offering software expertise in exchange for hardware components, for instance. Today's geopolitics point to a future in which self-reliance and security outshine economics.

"It will be the end user that foots the bill at the end of the day," warned Xu. "If fabs are increasing factories then chips will be more expensive and that will lead to higher costs for the whole industry." TSMC has already sent out letters to warn of price rises, while US sanctions have triggered "panic stockpiling" that could result in a future economic crisis, he said.

How much inventory Huawei has left is a question that still exercises the minds of analysts. Xu's hope is that alternatives not subject to US sanctions will emerge. "China is a huge market with enormous demand for chipsets," he said. "There will be companies moving forward willing to make the necessary investment to satisfy the needs of Huawei and similar Chinese companies. If that can be done, and if our inventory level can help Huawei to last until that time, that will help to address the problems and the challenges we face."

Yet most analysts think Huawei will start to feel the crunch in the next few months and say that developing a Chinese alternative to TSMC would take several years. The enterprise division is not growing quickly enough to compensate for a slowdown in the networks and devices units. And however good Huawei's software might be, the use of less sophisticated chips will probably hurt its competitiveness. This year could be a very difficult one indeed.