By Guest Author Nalini Kumar Muppala
Let’s continue from where we left off in the previous post, with an analysis of ST-Ericsson’s portfolio.
Even as it juggles the enormous task of integrating various teams, ST-Ericsson needs to be commended for continuing to deliver innovative products. Its Aero4228 solution, an RF transceiver for HSPA/EDGE, eliminates the need for expensive external filter components, thus saving money and space in the end design. The company also claims some industry firsts for sampling connectivity chips using 45-nm CMOS technology — this will be discussed in more detail below.
Combo chips are gaining momentum thanks to the cost and space savings they provide. As early as this January, at the consumer electronics trade show CES 2009, ST-Ericsson talked up combo chips as a big part of its strategy and roadmap. Prior to this month, the company had little to offer in the area of combo chips for connectivity — there was just the single-chip BT/FM part. The JV recently announced two single-chip parts: CG2900 and CW1100 using 45-nm process.
CG2900 is a single-chip GPS/BT/FM solution. It is not the first such combo chip: BCM2075 from Broadcom and BlueCore BC7830 from CSR offer similar functionality. However, ST-Ericsson is not just playing catch-up, as its sollution is the first to use 45-nm process technology. The power- and space-saving advantages of the CG2900 are due, in part, to its being a 45-nm part. The BCM2075, in contrast, is based on 65-nm. The CG2900 will not be in production until the second half of 2010, whereas competing parts should be seen in phones later this year. If the CG2900’s performance matches or exceeds the competition, the power and space savings will likely lead to an increased uptake of this part. This effort will further increase attach rates of GPS and BT.
CS1100 is a single-chip 45-nm WLAN part, another industry first. CS1100 claims to be a “ultra-low power IEEE 802.11n WLAN device”. The higher power consumption has been detrimental to attach rates of 802.11n in mobile devices. If the promised performance comes to fruition, it would mean a major breakthrough for ST-Ericsson. Again, competitors such as Broadcom have similar parts in production, albeit at 65 nm.
BT 3.0 achieves its high speed by riding on a WLAN PHY layer, inherently assuming that a WLAN part would be present in the device. Whether or not co-existence with a WLAN part is the best strategy for BT going forward is debatable, and one part of the discussion can be found here. There is no BT/GPS/WLAN/FM single-chip solution from anyone yet. Any design seeking to add all four connectivity options, at the time of writing, needs at least two discrete parts, just as ST-Ericsson recommends using CG2900 and CS1100 together. However, when a design chooses to provide just BT/WLAN/FM, competitors will steal design wins with a single-part solution. It appears that by not putting out a BT/WLAN/FM part, ST-Ericsson is betting on increasing attach rates for GPS.
The series ends tomorrow with a discussion of ST-Ericsson’s strategy and thoughts on the outlook for the company.