Radios could be a US$20+ billion product for Huawei in 2020 if orders come in. Very high sales are likely. China, Huawei’s primary market, plans 600,000-800,000 5G upgrades in 2020, some of which will involve multiple radios.

Huawei’s replacement of US parts is effective. Many designs require FPGA programmable chips, one of the few products where the US is still dominant. Ren implies that Huawei no longer requires FPGA’s. The only other important US-dominated product is radio frequency chips, where Huawei also is making progress.

Radios are second only to phones among Huawei’s main products. 5G is exploding in China and Korea. The rest of the world likely will follow.

Dell’oro analyst Stefan Pongratz and I had a long conversation about how hard it is to determine the right figures in 5G. Some announcements are the number of radios; others the number of bases. I’ve reported DT plans 300 for yearend. Many cells have radios for each of three sectors. DT’s 300 radios will be on 100 bases.

Financial and political issues affect what is reported. Huawei’s figure for 2019 radio shipments is higher than the numbers put out by the likely customers. The Chinese carriers are squeezed between government demands to accelerate 5G and investor’s desire for low capital spending.

China Mobile is requiring Huawei to lease/rent radios, which keeps the capex figure low. There is no information on how those units are counted, which might explain the difference between Huawei and carrier figures.

Companies issue figures that give a positive impression. They have many choices about what to report. Sales can be recorded at the time of shipment or delayed until the customers confirm acceptance. Nokia was hit by this in the first quarter of 2019. Nokia’s 5G had numerous problems when installed. Major customers refused to “accept” and pay for the equipment.

Everything about 5G is hyped. Separating facts from fantasy can be a challenge. First-rate analysts like Stefan closely review all the information from the companies. They look for outside datapoints that can confirm or refute claims.

For example, China’s claim of 344 million FTTH connections was wildly implausible. But I checked with the chip manufacturers, who confirmed to me they were shipping massive numbers of chips to China.