Increasingly, CES seems to be less about consumer electronics, and more about the big tectonic shifts in technology. Indeed, CES is no longer the Consumer Electronics Show, it is just CES. Of course, there were flexible screens (one you can even roll up). There were 8K TVs everywhere you looked. TVs that are as thin as a framed picture. Virtual reality goggles. Drones (but fewer this year, it seemed).

A big continuing trend is voice connection. Amazon's Alexa is everywhere, along with Google Voice, and even Siri now that Apple has opened it up so that it doesn't require a hardware chip to make use of the API. Cortana, Microsoft's entry, seemed less visible than last year. Perhaps that is part of Microsoft's increasing focus on cloud and business, rather than consumer gadgets and consumers.

One hall was almost entirely taken up with transportation. There seems to be an endless number of companies supplying speakers and amplifiers for very loud music. But the main thing flooding through the hall, almost like oxygen in the air, was autonomous driving.

One of the other big trends was 5G. In fact, no sooner had I written that I expected a marketing war to break out than...a marketing war broke out.

AT&T Fires the Opening Salvo

AT&T put out an update to their Android operating system specifically to change the network display so that advanced LTE (sometimes called LTE-A) is displayed as 5GE (and the E is tiny). That was the only change to that version of the operating system, no actual change in any functionality, even a trivial one. This was too easy to mock. As the Washington Post reported:

On Monday, T-Mobile tweeted a video showing a disembodied thumb applying a “9G” sticky note to one of its smartphones. The sticker covered up the label showing that the device was in fact connected to T-Mobile’s LTE network. “Didn’t realize it was this easy, brb updating,” the tweet said.

Our website renderer doesn't seem to like including a Twitter video directly, but you can see it here.

A couple of days later, in a keynote panel session, AT&T Communications CEO was presenting and tried to defend what they had done.

If I now occupy beachfront real estate in our competitors’ heads, that makes me smile,

Verizon got in on the game, publishing an open letter "When We Say 5G We Mean 5G." Except that Verizon started its 5G trials too early for the standard to have been complete, so when they say 5G, at least so far, they mean their proprietary version of the technology. To be fair, the trials are more intended to get early feedback about potential issues in the mmWave, which is new for 5G. Complaining about that is a bit like complaining that a testchip built to an early version of a memory standard isn't a production version—yes, but that's why we put the word "test" in testchip.

This all reminds me of 2G (CDMA and GSM). That was hyped as "digital" sound, so better than analog. It was the era of CDs, which had better sound than vinyl and cassette tape. Actually, 2G had worse sound, since it was compressed down to 16 Kbps, and then later 8 Kbps. It could fit more channels into the same bandwidth, so was cheaper, and eventually, that worked its way through to consumers. But the quality suffered from the voice compression. But it was undeniably "digital." Skype or a modern cell-phone isn't as good from a sound quality standard as the old POTS phones of most of the last century (and yes, POTS, which stands for Plain Old Telephone Service, really is the industry term). But POTS used 56 Kbps (64 Kbps outside the US) without any compression, and so would have been insanely expensive to use as the basis for 2G. But even today you still see phone company executives saying that digital sound quality was the reason to switch from 1G to 2G. Actually, there is only ever one reason to switch generations, better use of the radio spectrum, the ultimate fixed resource for wireless anything, and mobile phones in particular.

5G Value Proposition

The value proposition for every generation of mobile is always the same under the hood: trading off a higher performance of the silicon against more efficient use of the radio spectrum. Another way of looking at that tradeoff is that it requires higher performance digital signal processing (DSP) to handle more complicated radio encoding schemes. This results in higher bandwidth, lower costs (which eventually get passed through). At the level of a carrier, it allows them to build out more capacity, and give more customers (which might be "things" like cars rather than people) better communication service. However, that can be a big qualitative change. In wireline connectivity, broadband cable is just "more capacity" than dialup, but in another sense, it changed everything.

Like each generation before it, 5G both reuses the radio frequencies of the previous generation(s) and adds some new ones. The new ones are much higher frequency (20GHz) and are known as mmWave (pronounced millimeter-wave). The old ones are sometimes called sub-6GHz to distinguish them. Deployment of 5G is likely to be mostly in the sub-6GHz bands at first, with selective adding of mmWave inside buildings, and perhaps as a point-to-point way of providing broadband internet (in competition with cable and DSL).

mmWave

mmWave has some big positives and some big negatives. The big negative is that it is attenuated by air, so only operates over relatively short distances (1000ft or 300m), and can't go through walls (even tree leaves are a problem). The big positive is that there is a huge amount of bandwidth available. Speeds of over 100 Mb/s are possible (Gb/s in theory, but don't hold your breath). In the real world, your mileage may vary.

I actually expected the marketing war to be about blurring the distinction between what you get with a current 4G network, with what you get with mmWave. mmWave offers higher bandwidths of hundreds of megabits per second, but only over relatively short distances. So coverage will require a lot of small basestations. It's more like WiFi in some ways. The lower frequency bands are more like normal cell networks we are used to at 4G, although with somewhat higher transmission speeds. I still expect plenty of marketing that blurs this distinction, implying that you will get Gb/s to your cellphone as you drive around or are in your home. You won't.

CES

There is a phrase from Texas that someone is "all hat and no cattle." They talk a good story but don't actually do anything. 5G at CES was a bit like that. Everyone was talking about it, but nobody was doing anything about it yet. At least for real.

You could see some equipment. For example, the picture on the right is a Samsung 5G basestation. But nothing was running, and I suspect that the full electronics was not inside. There were a few handsets on view, but since all handsets look pretty much the same, I have no idea how real they were. There were even more fanciful approaches, such as 5G basestations on drones for use in sports stadiums. One of the reasons that the 5G standard was rushed through (it doesn't officially get signed off until April) is so that it would be ready for the 2020 Olympics in Japan. I'm sure there will be some 5G drones at the opening ceremony, if only for marketing. I did say that 5G is going to be all about marketing for a year or two, didn't I?

Mobile World Congress comes up at the end of February, and I fully expect that to be all-5G-all-the-time. Anyone with serious announcements, especially on the basestation side (which is nowhere close to consumer electronics), will be held for Barcelona. I will be there, so expect more Breakfast Bytes blog posts on the topic then.

What Will 5G Enable?

Higher bandwidth allows you to do more. Of course, some of that is just that, for example, your map will load faster on your phone. But a leap in bandwidth also allows things that wouldn't happen at the lower bandwidth. It is hard to imagine that we would access content on the net the way we do today if we were stuck with 56kbps dialup modems (which seemed really fast at the time...modems were 110bps in the days of 10 characters per second teletypes, so 56K was 500 times faster).

At some point, completely new applications become possible. For example, in 2005, Chad Hurley, Steve Chen, and Jawed Karim were the first people to get video right when they founded YouTube. Google acquired just 18 months later for what seemed a high price ($1.6B) but has to be one of the most successful acquisitions of the century. Don't forget, by the way, that was an era when people only watched videos on their PC. It was still a couple of years before the launch of the iPhone, let alone 3G and 4G that would make it even plausible to watch a video on your phone.

Perhaps an equivalent technology is already lurking, that we can't use on our phones, but can on our computers. One possibility is virtual reality (VR). Video, as in watching a movie, doesn't require very short latency (or even solidly reliable connectivity) since it uses buffering. But VR, beyond simply making VR equivalents of movies, requires interactivity. It is a cross between an online game and video. I remain somewhat of a skeptic about VR, although smarter people than me all say it is going to be the next huge thing. Delivering two 4K images with very low latency pretty much requires 5G (for both bandwidth and latency).

Autonomous driving is another area where 5G might help, but I remain a skeptic here too. All communication companies seem to say that 5G is essential for autonomous driving, but then the car companies pretty much never mention it. Perhaps in China, which has a more centralized approach to autonomous cars. But cars are simply not going to rely on a cellular connection, even if it is 5G, to decide if the light is red or that object is a child or not. As it happens, the keynote at DesignCon tomorrow is Robert Heath of UT Austin titled 5G for Vehicle-to-X Communications. Let's see if he changes my mind.

Mobile World Congress

As I said above, I expect Mobile World Congress in Barcelona to be where the 5G story is front and center. It is February 25th to 28th. Full details, including registration, are on the MWC website (not mwc.com by the way).

Sign up for Sunday Brunch, the weekly Breakfast Bytes email.