China just officially launched “commercial” 5G service, moving up the date from October 1st. The three Chinese giants previously disclosed plans for 90,000-150,000 cell sites in the fall. In comparison, British Telecom projects only about 2,000 cells in 2019. China soon also will pass the U.S., where most telcos have so few 5G sites active they won’t release the number.
Russia’s Sputnik, Britain’s Financial Times, and other Western media speculate without evidence that the U.S.-Huawei ban will hold back the Chinese. Very few key components can only be sourced in the U.S.
Fortunately, almost everything needed for a 5G network is available outside the U.S. or replaceable. 90% of the components used for Huawei’s 5G are not exclusive to U.S. vendors.
The most important problems are some radio frequency phone parts, primarily sourced from the U.S. Skyworks, Qorvo, Broadcom, and Qualcomm offer very small modules ideal for phones. Buying RF parts from Chinese suppliers today would make the phones slightly larger.
For memory chips, Huawei could go to Samsung, SK Hynix, and soon Yangtze Memory in China for replacements of the chips from Micron in the U.S. Huawei designed 5G chips, manufactured in Taiwan, match the capabilities of the Qualcomm and Samsung chips, the best in the world.
Huawei also designed its own chip for the 5G base stations, one of three in the world. It claims, “The 5G base station core chip Huawei Tianzhu achieves a base station size reduction of over 50%, weight reduction of 23%, power consumption savings of 21%, and installs in half the time required for 4G.
Huawei has a state of the art network processor, with the processing power for the routers and switches. It even makes optical components, one of the most challenging technologies.
Huawei has delivered 100,000 base stations, including to LG in Korea, British Telecom, and Etisalat in the UAE. I see no obstacle to rapidly ramp up production into the hundreds of thousands.
ZTE is also ready to produce high volumes of base stations. A third vendor, Datang, is less known outside China. Datang played a historic role in developing TD-SCDMA, China’s alternative to Western standards. TDD is an essential part of 5G.
Datang had some unfortunate years and fell behind in 4G. In 2018, it was merged with Fiberhome in Wuhan, a world-class optical producer. The company owns foundation patents for 3G, 4G, and 5G.
It has ambitious plans to become the world’s sixth major supplier of 5G systems. The company recently brought 5G systems to Germany’s important ANGA COM and has offices in 50 countries.
If Huawei is unable to produce phones in sufficient volume, Oppo, Vivo, Xiaomi, ZTE and Lenovo could easily pick up the slack. All are shipping in volume.
FPGAs (Field Programmable Gate Arrays) are the other important class of chips primarily from the U.S. An FPGA combines numerous functions into a single chip. They are cheaper than custom chips when fewer than ~tens of thousands are required.
Huawei is producing enough units it already is probably designing custom chips to replace FPGAs. If not, base stations, routers, and the other components using FPGAs have enough room to substitute half a dozen chips if necessary.
The limiting factor on China’s 5G growth will probably be the time required to train the engineers.