View from Washington: How do you solve a problem like Huawei?
Image credit: Pixabay
Concerns about collateral damage from Washington's face-off with the Chinese telecoms giant are growing.
The US Department of Justice’s unsealed indictments against Chinese telecoms giant Huawei, its CFO Meng Wanzhou “and others” have provoked a range of reactions from western high technology.
On one side, there is satisfaction that the Trump administration is following through on its rhetoric about China’s lax – many would claim still larcenous – attitude to foreign intellectual property (IP).
The IP issue has been a Silicon Valley bugbear for decades. However, it appears to have been accentuated now by the allegation in one indictment that Huawei encouraged staff to steal from rivals and offered bonuses to those who did (page 19, paragraph 47 here).
Let’s set aside those long-standing concerns, because - alongside unquestionably serious allegations that also cover sanctions-busting - another issue lurks in the background: can the global 5G market mature outside China, as it needs to, without Huawei? It is the technology’s largest and arguably most advanced supplier.
Following the indictments, questions that have been asked ever since Washington began to ratchet up its attacks on the Chinese conglomerate have taken on greater urgency.
They reach beyond infrastructure. The Internet of Things (IoT), artificial intelligence (AI), machine learning and machine-to-machine (M2M) communication are seen as fundamental to returns on 5G investments. They justify the capital cost for a technology that, according to Deloitte, will not ramp among consumers much faster than 4G.
Many M2M applications will indeed operate over unlicensed, short-range wireless networks (Wi-Fi, Zigbee, Bluetooth etc) and pass data over wired broadband. 5G is nevertheless seen as a source of extra capacity (Cisco forecasts that M2M connections will account for almost 3ZB [zettabytes] of all Internet traffic by 2022) and as fundamental to applications in areas such as autonomous driving, e-medicine and agriculture because of speed, lower latency and, eventually, ubiquity.
Analysts reckon that Huawei today has a healthy technological lead over its main 5G infrastructure rivals, Nokia and Ericsson. It has invested $1.4bn (£1.1bn) in research and development during the last decade and participated widely in the standardisation process at the Third Generation Partnership Project (3GPP).
Similarly, the company’s willingness to compete aggressively on hardware prices is thought to have made some 5G proposals economically viable.
In that context, here are five of the key questions those who feel uneasy about the Washington-Huawei/China standoff are asking, but which have not got as much of an airing as the allegations surrounding IP and also spyware.
Like many such questions, they tend to suggest worst possible scenarios. Those are unlikely to come about in full. The aim instead is illustrate why uncertainty is spreading and why, in the words of Vodafone CEO Nick Reed, this crisis is being addressed at a “too simplistic level”.
Could ‘No Huawei’ mean ‘No 5G’?
In global terms, no.
US carriers have already excluded Huawei. China will move ahead as originally planned. Most of Europe seems prepared with, for example, Vodafone ‘pushing pause’ on supplies from the company last week, while BT/EE has excluded Huawei from bidding on its core 5G network – both regardless of any specific instruction from Brexit-bedevilled Westminster.
In Australia, one putative carrier, the wired broadband operator TPG, has just abandoned plans to enter 5G because the Australian government has followed the US lead (and encouragement) in banning Huawei from infrastructure.
“In light of the government’s announcement in late August 2018, [the] upgrade path has now been blocked,” TPG said. “It does not make commercial sense to invest further shareholder funds.”
Will others reach the same conclusion and - far more likely than no 5G at all - could the rollout therefore be slower, geographically patchy and have fewer competitors than expected? Nobody knows, but TPG is a warning sign of that rising uncertainty.
Could the rest of the 5G supply chain support ‘No Huawei’?
As at TPG, the lower cost of Huawei equipment featured in many 5G operators’ original business assumptions. If they find themselves forced to turn to (and can afford) other sources – as is already the case in Japan and New Zealand, as well as Australia and the US – will Huawei’s rivals be able to fill the gap anyway?
According to IHS Markit, global infrastructure market shares in 2017 had Huawei at 28 per cent, Ericsson at 27 per cent, Nokia at 23 per cent, ZTE at 13 per cent and Samsung at 3 per cent. These figures do not fully reflect Huawei’s overwhelming leadership in China. Moreover, Huawei’s numbers skew more heavily toward newer-generation technologies. A completely Huawei-free West would have to find a lot of 5G equipment from somewhere else.
Rollout projections for 5G in major international markets that spread between now and the end of 2020 might suggest there is time for rival vendors to scale, but you need to remember that cellular builds precede launches by months, sometimes years. These are literally national projects with extremely complex supply chains.
Switching suppliers might be possible, but schedules could become very tight and notably, in some cases, 5G will require a bigger build than 4G because of distance and object limitations on the signal if it has to be delivered at higher frequencies. Or they could be delayed?
Could ‘No Huawei’ seriously shrink VC funding and ROI projections?
Could a slowed 5G rollout demand sacrifices to the most tyrannical force of the digital age: time-to-market? Such delays would apply to all players, but the immediate issue here is less about competition than investment.
How quickly have venture capitalists been expecting to see returns on their 5G-related bets and what might their response be if these look likely to arrive late? Will they become more risk-averse and/or slow down their funding programmes?
VCs generally avoid communications infrastructure and favour applications and other products that sit on it precisely because they get a faster ROI from the latter. Another important nuance, then, is that many of the companies preparing products that will exploit 5G are at or are moving towards second- and third-stage funding. The sums these startups want are getting bigger.
If one takes AI data analytics as an example, one trend there is seeing companies add silicon development to their initial focus on software and algorithms. They are taking on these NREs and accelerating their burn rates because their offerings are becoming more complex and stressing generic GPU-based AI platforms. ASIC is coming to mean “algorithm-specific” as well as “application-specific”. Kerching!
How long to process Huawei’s own ‘No’?
The US Department of Justice has a history for inserting itself in technology markets late in the day - and having processes that take too long to deliver ‘justice’.
Remember the first browser war between Microsoft and Netscape? Netscape was forced to sell itself to AOL well before any legal anti-trust judgement was handed down. Its management had little option, even though its argument about the damage it suffered by the bundling of Internet Explorer with Windows was strong.
Anyone care to bet on how long the two Huawei indictments might take to go through court, adding all the while to market uncertainty? No? Thought not.
Could ‘No Huawei’ in the West hand both the 5G and AI crowns to China?
This is probably the most nebulous but potentially the most significant question.
The ‘Made in China 2025’ plan has both AI and 5G at its core. Beijing has identified these as two areas where it can not only catch up with but leapfrog technological rivals in the West.
In 5G, China is reckoned to have helped define about 40 per cent of the standards at 3GPP with Huawei leading the effort. It was very much a bystander when earlier standards were set.
Then in AI, Chinese specialists received $4.5bn in investment between 2012 and 2017, and the sector is today estimated to be accounting for 60 per cent of AI funding globally.
In that regard, a stuttering 5G rollout elsewhere might suit Beijing well. Huawei will continue to build out its ‘local’ network and the Chinese applications running on it will have a large market in which to mature before others can really start to scale.
Moreover – and irony of ironies – if Huawei is restricted from selling hardware to much of the rest of the world, China could accelerate its domestic 5G build not just to compensate, but also to inject money into its economy as a major slowdown threatens. See: rural broadband – quantitative easing – Obama – passim.
Of course, there is the caveat that if you don’t trust Huawei, having it in Western networks might also be a case of roasting your chickens for the fox. Nobody said any of these issues are easy to resolve or can be entirely resolved right now.
There is also that rather persistent and overarching factor: respecting the rule of law.
Still, there you have it. IP security is a huge issue, as is the controversy around spyware, though it’s hardly touched on here. As Nick Read said, this is an issue that, even taking those factors into account, is being addressed at a “too simplistic level”.
The next couple of days of trade talks in Washington between the teams of US President Donald Trump and Chinese Vice Premier Liu He might brighten the mood and offer a few pointers. Maybe this is just the Donald being a blowhard again.
All you can say for sure, though, is that these known unknowns are at least good fodder for columnists.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.