A week of shocks – industry mourns ARM; Ericsson CEO under pressure
As Softbank buys ARM and Facebook tests yet more advanced wireless networks, the world is changing too quickly for traditional spectrum and business models
This week started with a bang – we UK dwellers woke on Monday morning to the news that the country’s shiniest hi-tech star, ARM, was being sold to Softbank.
That shock event continued to dominate the debate for the rest of the week. Many parties weighed in. The UK government welcomed it as a sign that the UK was still ‘open for business’ despite Brexit, even though the timing of the bid blatantly related to the stricken pound, and other fire-sale deals may follow.
Softbank gloated, as well it might. Its chairman, Masayoshi Son, is taking a huge gamble on an ambitious vision to create a platform for the next generation of internet services, especially the IoT, which it can drive from silicon to networks, devices to applications. Whether it will succeed is open to serious question, but if it does, the jewels ARM brings in IoT processing, security and the mBED OS will have helped – even if Softbank destroys the ARM business as we know it, in the process.
This seems almost inevitable. ARM’s raison d’etre has been to be neutral to all its many competing licensees, which is why persistent rumors over the years, that Intel or Apple would buy it, always came to nothing. Softbank is not as bad as that, because it is not a chip player, and for the near term it has promised to keep business running as usual at its new subsidiary (assuming the deal clears). For the coming generation of chips, anyway, the licensees will have nowhere else to go.
But there would be no point in buying ARM if Softbank did not change the model and harness future developments for its own agenda – even in the company’s ragbag of businesses, a processor IP play would not make sense in its own right, despite the nice boost to profits. So while some Wall Street analysts were concerned about the deal because they feared it might leave Softbank with less free cash to bail out Sprint, they should really have been more worried about the medium term effects on key US ARM customers like Qualcomm, whose future architecture strategies will now have to be devised with more unknown quantities than investors would like.
Away from M&A, the week’s themes were customary ones in wireless – the quarterly round of earnings results, and spectrum issues.
Most notable on the earnings front were signs that the smartphone market is more robust than pessimists had expected, though nobody is yet changing forecasts that the full year will see sales rise by only about 3% (the first year below double figures). The main, and predictable, trend is that the big guns are marsalling their resources to ensure they outperform that 3% curve and consolidate more of the higher value market among themselves. Samsung’s smartphone recovery and a very strong quarter from Qualcomm – which was down to better-than-expected mobile chip sales as well as IP gains in China – were highlights.
The other big earnings headlines both concerned market giants under pressure. Intel shocked shareholders with considerably worse-than-expected sales in its critical server business, which it urgently needs to sustain growth to offset PC decline, smartphone failure and the immature state of other potential areas of new business, such as IoT, Cloud-RAN and the newly acquired Altera FPGAs.
And Ericsson’s CEO, Hans Vestberg, is under pressure to resign after the Swedish firm turned in a poor quarter, with a revenue drop of 11% year-on-year, and said it would have to double its cost-cutting targets to cope.
Spectrum, of course, is always in the wireless news, though nothing as dramatic as the FCC’s groundbreaking decision on rules for millimeter wave 5G bands last week. That was followed by the White House issuing plans for an Advanced Wireless Research Initiative, which will help create the technologies to run in that high frequency spectrum, among other things – but felt rather belated. Its objectives, and the language of its launch, sounded very like the 5G public-private R&D programs which the European Union, China, Korea and others were kicking off several years ago. The FCC also announced the long list of bidders for the 600 MHz auction coming up soon.
With all the talk of 5G, it’s easy to forget that, in many parts of the world, LTE spectrum is still to be allocated. The GSMA chided the European Union for being slow to get into line with 700 MHz, but its real concern was reserved for Egypt, where 4G spectrum is only just being auctioned, and it fears there will be too little for operators to launch competitive new services. Of course, India’s latest, and massive, 4G auction looms, though several players have been deterred by high reserve prices – including Telenor’s Indian unit, which is sitting out the auction and may exit the market altogether, as Russia’s Sistema already did.
And then there’s Facebook. While regulators, operators and the GSMA bicker about auction processes and small slices of traditional spectrum, the social media company is becoming a disruptive force by showing what a future network really might look like. We had the Terragraph urban mesh and Aries massive MIMO antennas earlier in the year; the OpenCellular open source small cell concept last week; and this week the firm announced the first full test flight of its Aries high altitude drone, which aims to enable connectivity in the remotest areas.
Plus its Connectivity Lab demonstrated another conceptual approach to delivering internet services to far-flung communities. This uses free space optics technology – which may be set for its latest burst of hype – but with fluorescent materials instead of traditional optics. Combined with conventional telecoms this was able to deliver 2Gbps speeds with OFDM modulation in a 100 MHz channel. Facebook researchers are now interested in developing new materials which would enable the system to work in infrared spectrum, which would boost speeds beyond 10Gbps.
Whether or not this particular piece of R&D proves commercially viable remains to be seen, but it does show the new players which are now driving some of the wireless industry’s R&D agenda. With new technology breakthroughs come new ways of thinking, and just like ARM’s licensees, the mobile operators and vendors are going to have to adapt quickly to live in this new world.