China Mobile reports 31 million 5G "contracts" at the end of March. That's an increase of 16 million in one month. If that pace continues, by the end of 2020, over 170 million of China Mobile's 946 million wireless customers will have 5G contracts. It will reach 350 million in 2021.

China Telecom reports 17 million. China Unicom didn't offer a figure, but typically is about half the China Telecom figure. The total is over 50 million.

But probably only half are actually connected to 5G. A second source reports only 6 million 5G mobile phones were sold in China in March. Even adding some routers for fixed wireless, many of the 5G contract holders were still using 4G phones.  One estimate is that only 20 million are actually connected.

The gap will probably narrow. Phones are just becoming widely available at prices between US$285 & $450.

150 million actual connections by the end of 2020 are likely.

3GPP calls the coming release of 5G standards "Ultra-Reliable Low Latency Communication" (URLLC). However, key members of 3GPP strongly disagree. Jason S. Boswell, Ericsson's head of network security for North  America, cautions, "[5G] will create new mobile security challenges. There will be broader attack surfaces, more devices, and greater traffic."  

Primary 5G safety is not adequate. Instead, those needing certainty will have to pay extra for "tailored security solutions such as 'network slicing.'" These enhancements will be necessary, although not a guarantee. 

Boswell sees a problem. "We must have networks that are trustworthy, resilient, and secure by design – all on day one." Given that 50 million people are already connected to 5G that will be difficult.

The standards are not yet ready.

Some politicians think keeping Huawei out of U.S. networks ensures security. I believe that does very little. The Chinese government does not need to compromise Huawei because there are so many other ways to crack the systems. Former FCC Chair Tom Wheeler and his security lead, Rear Admiral David Simpson note

"Never have the essential networks and services that define our lives, our economy, and our national security had so many participants, each reliant on the other—and none of which have the final responsibility for cybersecurity,"

5G networks will be vastly more complicated, with far more points of attack. Software-defined networks are harder to protect than traditional, known hardware. Control - and vulnerability - is more distributed. Running on commodity hardware and software offers more ways in. 

That 5G is not secure is well known. Congress was told that by Ericsson’s Lynn Starr, Barbara Baffer, and Jason Boswell, presumably informed by the engineers building about half of U.S. 5G networks.

"5G is ultra-reliable" is a lie.

CEO Dirk Wössner's "5G" claims are now meaningless. DT is now using DSS and calling connections "5G." Since the actual speed is about the same as 4G, that proves the industry definition of "5G' is now meaningless.

It's so slow - and the latency so high - DT refuses to provide figures. Verizon President Ronan Dunne has a similar offering, but explains, "The lower down the spectrum tiers you go, the more that will approximate to a good 4G service." That's what DT is now selling, a good low-band 4G service.

Until now, DT had said it would bring 5G to 20% of Germany by the end of 2021. It didn't consider low-band "5G," even though it uses 5G NR software.

DT has changed its actual network and capex plans. It changed what it considered 5G to the 3GPP marketing dream or anything that runs "5G NR" software. The software has only a minimum effect on the performance of below 2.5 GHz - 19% or probably less. DT is providing no figures, but its new 5G is going to be slower than some available 4G.

Minister Miao Wei demanded, "Accelerate" and the telcos are delivering. 198,000 cells had been upgraded at the end of March. The year-end goal, 600,000, has been reduced to 500,000. Ericsson and Huawei have confirmed the equipment is on the way and the buildout should hit the new target easily.

The 50 million "5G contracts" figure must be examined closely. Many reporters assume that means 50 million people are using 5G. I initially did. 

However, I discovered in March that the Chinese telcos were selling more 5G contracts than the phone makers seemed to ship.

With 5G phones as low as $285 in China (Xiaomi), it was inevitable that low prices would expand worldwide. Giant TCL just announced a US$400/Euro 400 6.5" smartphone with a Qualcomm Snapdragon 765G processor and four cameras on the rear.

Chinese company TCL has been selling under the name "Alcatel" in the west. Its sales were over 60 million phones in 2016 but it has fallen behind. It intends to deliver between June and September. The Chinese factories, include Foxconn, are close to full production.

Oppo, Vivo, Xiaomi, and ZTE are similar phones in China at $285-$400. Samsung and Huawei will soon follow. As soon as these prices come West, demand for 5G will expand rapidly.

Unless everyone decides to buy the new iPhone SE at a similar $400.

VTEL Wireless can now deliver 100 Mbps to 80% of Rutland, Vermont. With help from Ericsson, this took less than 2 weeks. The company is contributing modems and service, the school district is providing Chromebooks. The kids are learning.

~8 million U.S. homes do not have cable or a decent landline. Most can be reached from existing towers and upgraded in weeks. It should be done, whether by T-Mobile, Verizon, AT&T, or regional carriers. Ajit Pai should make this so.

Massive MIMO, invented at Bell Labs in 2010 by Tom Marzetta, uses 32-128 small antennas to deliver multiple data streams where the local demand is highest, 

Below 1800 MHz, 5G has almost no performance advantage over today's 4G LTE. For a major power grid running at 450 MHz, PGE Systemy chose Nokia 4G, resulting in a major saving. 5G is overkill for almost all IoT applications. 5G's NB-IoT is almost exactly the same as 4G NB-IoT.

The key exception is when connecting an enormous number of devices. 5G can address a million devices from a single tower. So far, I haven't found any application that needs more than 4G can handle. It's not clear whether intense concentrations of cars, all on one network, will need that. Perhaps when 5G controls munitions, the U.S. will need that capacity on the battlefield. The U.S. military is massively funding 5G research, especially on how to manage spectrum in a presumably crowded environment.

Nokia virtually "bet its future" on corporate 4G & 5G networks.