IDCC mentioned.
Musical Chairs in RF Chip Market
By John Donovan
Will StraussGuest Blog by Will Strauss, President & Principal Analyst
Forward Concepts
Infineon Bidding War?
It is clear that Infineon has put its cellular handset chip business on the block and rumors are rife on who is bidding and who the likely purchaser will be. In mid-May the Financial Times Deutschland published an article stating that Intel is interested in acquiring the cellular chip properties of Infineon Technologies AG, which provides the cellular modem and RF transceivers for all iPhones and the iPad. Apparently, Intel was initially rebuffed. However, in mid-June, Infineon announced that it had engaged investment banker JP Morgan to study the possibly of selling their cellphone chip business. I guess Infineon realized that they could make a nice profit through such a sale and they could concentrate on the more lucrative automotive market.
By engaging an investment banker, Infineon is certainly interested in getting multiple bids for the cellular business and realizing the highest price possible. Rumors abound that in addition to Intel, Samsung, Broadcom and even Apple are among those in talks with Infineon. I think we can forget the Apple rumor, since even Nokia has decided it doesn’t want to be in the chip design business anymore (see Renesas, below). The asking price has been reported as about Euro 1.5 billion ($1.9 billion), so let’s examine the possibilities:
–Broadcom: The company currently has a very small share of the 3G cellphone chip market, although it has done rather better in the 2.5G market (with Nokia as a major customer) and is a major supplier of ancillary cellphone chips, which include Bluetooth, Wi-Fi and GPS. In the past, Broadcom has paired its 3G baseband chips with Infineon’s worthy 3G RF transceivers, so the two companies are well-acquainted. Infineon’s significant sales of both 3G transceivers and basebands to Apple (perhaps 40% of Infineon’s cellphone chip shipments) would be a major boost to Broadcom’s market penetration. So the only question would be in Broadcom’s possible commitment to such a large investment.
–Samsung: The #2 semiconductor company in the world would like to be #1, displacing Intel. The cellphone market would enable it to quickly expand its revenues in a non-memory market that could eventually rival the PC market. The handset side of Samsung has been a customer for Infineon’s 3G RF transceivers, so that could be additional incentive to acquire Infineon’s cellphone chip business. Samsung could certainly afford the rumored purchase price. However, would LG, Nokia, and even Apple want to buy cellphone chips from a rival cellphone vendor?
–Intel: the #1 semiconductor company has always coveted the communications market, and they realize that wireless could eventually rival the PC market. Moreover, Intel has hopped in bed with Nokia on a number of wireless things, like licensing Nokia’s 3G modem technology and jointly developing MeeGo, the Linux-based operating system that Nokia plans to use in its future E-series handsets for the business market. It appears that the Nokia-Intel modem licensing did not include 3G/HSPA RF transceiver design, something that Infineon is good at (and few others are). Unlike Samsung, Intel would not be seen as a threat by cellphone vendors, and everything points to Intel’s strong wireless ambitions, so we pick Intel as the likely buyer.
Fallout from Infineon Sale?
If Intel purchases the Infineon cellphone chip business what will happen to Comneon, the Infineon subsidiary that licenses 3G/HSPA/LTE software stacks? Intel has never entertained the licensing model as part of its chip business and would likely shutter the Comneon operation, just as Motorola did after they purchased TTPCom’s 3G licensing business. TTPCom’s 3G stack lives on (albeit with much improvement) as a component of Icera Semiconductor’s modems, MediaTek’s new WCDMA modems (via its Analog Devices cellular chip acquisition) and Marvell’s 3G modems employed in a number of RIM BlackBerry models (BTW, is the plural of BlackBerry® BlackBerries?).
There are a number of smallish LTE baseband startups (mostly morphed WiMAX baseband suppliers) that are aiming at the USB data dongle modem market. And, of course, the data dongle market is currently the only LTE terminal market, and will be for perhaps another year or two. For LTE cellphones, fallback to HSPA/3G/2G will be required, since the world will not instantly change over to LTE. Faced with investing an estimated $300 million and many man-years to develop a complete 3G stack, it is likely the only possibility for the new LTE modem houses is to license the technology. The possible sources will likely be narrowed to two: InterDigital Communications (which was a joint developer with Infineon of the Comneon stack) and NTT DoCoMo. DoCoMo announced last week that it had licensed its LTE stack to MediaTek to augment its current 3G/2G stack.
Renesas Rising
At the Mobile World Congress last February, there were rumors that Renesas was soon to get a 3G/HSPA/LTE baseband modem design win at Nokia. When I met with Renesas executives at the Congress, they readily admitted that they were beefing up their Munich operations to better serve Nokia. But the emphasis was on Renesas’ RF transceiver and RF power amplifier chips and every time I asked about a baseband chip, they answered the question, but like politicians, never answering the question that was asked. So, I knew that something was underway, and it’s now clear that Renesas had licensed the Nokia HSPA+/LTE modem IP last year, and Nokia must have liked the progress made by Renesas.
In early July, Nokia and Renesas jointly announced that Renesas was to acquire Nokia’s wireless modem business for $200 million. That acquisition involves transferring 1,100 Nokia R&D professionals who are mostly located in Finland, India, Denmark and the UK (a no-pain layoff?). It appears that Nokia found it too expensive to continue developing its own modem designs and felt that it was more economical to acquire the modem chips from outside sources. The deal is set to close later this year. Apparently, Nokia will continue developing 3G/HSPA+/LTE software, which it will share with its modem licensees, which now also include Broadcom and Intel (who are not presently shipping cellphone modem chips based on the Nokia IP). It’s clear that Nokia will continue purchasing modem chips from Qualcomm, ST-Ericsson, Infineon and Broadcom, so they can get competitive pricing (at least in non-CDMA modems).
You’ll recall that Renesas merged with NEC Electronics’ chip business a few months ago, and collectively, the two only had about a 2% market share in the 3G modem market, mostly based on NTT DoCoMo IP, and selling primarily to Japanese cellphone companies. Now, the Nokia acquisition will enable them to expand out of Japan and be able to provide competitive modem chips on a global basis. Although the “original” Renesas has competitive RF chips, it’s unclear whether the company will continue fielding its original application processors based on SH-DSP or will use the NEC-heritage EMMA chips based on ARM Cortex-A8 and A9 designs; likely the latter, since Renesas had also licensed the Cortex-A8 technology.
Forward Concepts
Tempe, AZ
www.fwdconcepts.com
Email: wis@fwdconcepts.com
