Where is GPS going in the handset market?

Implications: GPS integration into mobile handsets will significantly increase the unit shipments for GPS. However, the form that GPS takes will determine the effective ASP of the GPS technology. Both the unit shipments and the effective ASP are key elements in determining the future of GPS technology suppliers.

Analysis: For several years now Qualcomm has integrated GPS technology into their CDMA chipsets, combining the GPS baseband functions into the CDMA baseband IC and the GPS RF circuit into the CDMA RFIC. The baseband integration is typically the most cost effective and power efficient due to the fact that the cellular baseband chips are developed on process technologies that are 1 or 2 generations ahead of the stand-alone GPS chips. Such integration further improves cost by eliminating a significant number of I/O drivers and pads from the GPS function, and by allowing the GPS to share memory resources already in place for the cellular baseband IC. The efficiency of integrating the GPS RF circuit into the CDMA RFIC is a bit less clear as the circuitry eliminated is minimal and the process migration is less advanced.

This GPS integration path appears to be the direction some companies expect cell phones to travel. The recent acquisitions of GPS IC companies by Broadcom (Global Locate) and NXP (Glonav) suggest that at least these two silicon suppliers are betting on that direction. Both TI and STMicro also have captive GPS IC technology, TI being a supplier of GPS for some of Nokia's GPS-enabled handsets.

With Qualcomm already integrating GPS into some of its WCDMA chipsets, it seems reasonable that other suppliers hoping to compete in the WCDMA market will offer similar levels of integration. As this trend occurs, the value (or effective ASP) of the GPS function becomes less clear. However from a pure cost stand-point it can be calculated as the percent of silicon area occupied by the GPS function times the ASP of the baseband IC. This calculated value is much lower than that of a stand-alone GPS IC and is likely to have an impact on the ASP that a stand-alone GPS might command.

An alternate integration path (one that some cell phone manufacturers are supporting) is the multi-function radio IC. In this case, the various radio functions in the handset (Bluetooth, WiFi, GPS, etc.) are integrated into a single device. Atheros appears to be banking on this model with its recent purchase of GPS chip vendor uNav Microelectronics. SiRF has also made comments about their efforts on multi-function radio ICs and over the past two years have acquired technology to support this goal.

The multi-funciton radio IC concept has certain efficiencies due to the similarity of process technology required for each radio function, and the possibility that certain RF circuits may be shared by more than one radio function. The integration is also likely to reduce packaging and test costs, as well. However this model still results in a stand-alone chip and the logistics associated with it.

Of these two GPS integration models, the chipset integration (ala Qualcomm) has the early lead and significant volume behind it. If one assumes that the evolution of cellular handsets follows the evolution of the PC, expect the chipset integration to win a dominant share over the multi-function radio IC. That being the case, companies focused on the multi-function radio IC model might be well served by hedging their bets.