About a dozen market research forecasts predict the number of 5G smartphones will range between half and one billion units by 2023, rising faster than the adoption rate of 4G LTE smartphones. However, 5G devices are best seen segmented into three main segments.

In the next three-to-five years, the main segment will be for sub-6 GHz handsets, representing an evolution of the 4G telecom standard. Early high volumes will be in the new 3.5-4.5 GHz frequency range.

Sub-6GHz 5G technology is critical since it adds a large frequency spectrum with wide bandwidth to the current heavily saturated 4G and 3G cellular bands, mainly concentrated below the 2.1-GHz frequency. Interestingly, the current 4G technology will continue to evolve and complement sub-6GHz 5G for a variety of reasons I won’t discuss here.

A second, moredisruptive and revolutionary segment is in millimeter wave (mmwave) 5G, at present mostly at 28 and 39 GHz. The 5G mmwave capability in smartphones will be added on top of sub-6-GHz 5G. By the year 2025, only a third of 5G smartphones will likely have mmwave capability, giving components for the sub-6 segment a 3:1 edge.

5G handsets Petrov

Handsets supporting mmwave bands will only make up a third of 5G smartphones by 2025. (Source: Petrov Group)

A separate 5G IoT segment addresses data transfer from a very large number of end devices at below gigahertz frequencies. The standards and protocols for 5G IoT are not defined yet, in part because initial optimistic expectations of cellular IoT in 4G have not been met despite big efforts, especially in China.

It’s important to note that in the next few years many smartphones will claim to have 5G capabilities that a network operator can enable with a simple software upgrade of its network. However, the sub-6-GHz and mmwave segments are distinct and significant for several reasons, including regional differences, potential for supply chain disruption and for positioning of smartphone and infrastructure vendors.

In the near term mmwave 5G is important primarily in the U.S. market, as well as in the much smaller Japan and South Korea markets. It is aggressively promoted by network operators who pushed for municipal installation permits for base stations in a less time consuming and less expensive manner.

Actual use cases for mmwave 5G are somewhat unconvincing and uncertain due to cost and performance challenges. The much larger sub-6GHz market is expected to blossom in Asia, mostly in China. Thus, it could be dominated by Huawei’s wireless infrastructure equipment.

The two segments are significantly different in their likely effect on supply chains in RF front end (RFFE) modules. The sub-6GHz segment likely will extend the current dominance of RFFE module vendors Broadcom, Skyworks, Qorvo, Murata and the more recent fabless entrant, Qualcomm.

The mmwave segment brings challenges in cost, low power, new materials, packaging, and testing. In particular, mmwave transmitters need to be very close to antenna arrays, preferably in the same system-in-package. Thus, they are likely to disrupt the established supply chain, giving an advantage to modem suppliers such as Qualcomm, Samsung, HiSilicon, MediaTek, and Unisoc.

Huawei/HiSilicon is effectively prevented from competing in the U.S. 5G market today. That’s giving Samsung a big incentive to compete in 5G segments against Apple and its supplier Qualcomm.

A high priority for Samsung is to build up its volumes in foundry wafers as well as IC packages (note its investments in panel-level fan-out capability) in order to compete against TSMC. Samsung already has reclassified support of its smartphone group as part of its foundry business.

Samsung also stated a goal to become the third largest global supplier of wireless infrastructure equipment. In the next few years, it will be very active in achieving its corporate strategy goals.

Overall, market segmentation for 5G smartphones will continue to evolve. And 5G has major implications beyond smartphones for a large number of other markets and industries.

-Boris Petrov is managing partner of the Petrov Group. He spent the last ten years leading market intelligence and strategic marketing at the SCL/JCET group and previously held positions at Intel, Zilog, Siemens, McKinsey, Boston Consulting, and Booz-Allen.