Small cells

Broadcom vs. Qualcomm Goes Femto

The long-standing battle between two of the wireless silicon market's big guns, Broadcom Corp. (Nasdaq: BRCM) and Qualcomm Inc. (Nasdaq: QCOM), is about to enter the femtocell arena now that Broadcom has announced its intention to acquire Israeli femto chipset startup Percello Ltd. (See Broadcom to Buy Femto Chip Startup for $86M.)

The clash of these two titans is not likely to change the femtocell scene overnight; however, it could lead to a new but related battle, as the duo's rivalry could now spill over into the broadband home gateway sector. (See CTIA 2010: Qualcomm Snags First Femto Wins, Qualcomm Dishes New Femto Details, and RadioFrame's Mystery Femto Buyer.)

The possibility of adding femtocell capabilities to home gateways is one of the reasons why Broadcom likes the small home base station technology, and why it decided to buy Percello for $86 million. (See Broadcom Spotted at Femto Forum.)

"Femtos, or small cells, are potentially disruptive technologies and synergistic with what Broadcom does with broadband, backhaul, and WiFi," said Greg Fischer, VP and general manager of Broadcom's Broadband Carrier Access line of business, in an interview with Light Reading Mobile.

The acquisition also means Broadcom will square off with its rival Qualcomm, as Fischer noted that the acquisition "brings us head-to-head" with the San Diego-based chipmaker.

Qualcomm and Broadcom both have their eyes on integrated femtocells, according to Heavy Reading senior analyst Gabriel Brown.

"Broadcom builds so many of the other components in a home gateway," says Brown. "It can possibly start creating a better optimized platform or reference design, integrating femtocell capability into home gateway platforms. This is what Qualcomm is proposing in the long term with Snapdragon."

Of course, Broadcom and Qualcomm aren’t the only femto chips players in town. Fischer noted that the other "key competitor" is Picochip , the UK-based femto chipset pioneer that claims to have 80 percent of the femto market and expects to ship 1 million chips this year. picoChip's products are used in commercial femto services at AT&T Inc. (NYSE: T) and Vodafone UK , for example. (See picoChip Scores $20M, Ships 1M Chips , PicoChip Bags $20M, Plans IPO, picoChip Intros Low-Cost Femto Design, Marvell Man, and picoChip Expands in India.)

Integration isn't everything
While Broadcom's Fischer said the company is already hearing interest from operators about a variety of integration configurations involving Ethernet, WiFi, DSL, and femtocells, he indicated that those various combinations would be realized once femtocells reach a certain volume in the market. But integration isn't necessarily the key to increasing the market volumes.

"I don’t think that necessarily reducing the cost through integration would make the technology prolific," he said. "Operators have other challenges yet in terms of figuring out their strategies as well as the technical challenges. There are guys leading [AT&T, Vodafone, SoftBank Corp. ] -- other carriers are watching how those deployments go."

"We have a long-term view on this," added Fischer. "[The Percello acquisition] is a very good start for us."

So what other silicon suppliers might be looking to make a move on femtocells? Companies to keep an eye on are Freescale Semiconductor Inc. , and Cavium Inc. (Nasdaq: CAVM), as well as Texas Instruments Inc. (NYSE: TXN), which introduced a digital signal processor for femtos in late 2008. (See TI Dives Into Femtos.)

— Michelle Donegan, European Editor, Light Reading Mobile

KCDay 12/5/2012 | 4:19:51 PM
re: Broadcom vs. Qualcomm Goes Femto

At Ubiquisys we reckon that Percello has been consistently underestimated, both for the quality of its products (performance, price and energy footprint) and for the market share it has established.  Broadcom has made an astute acquisition. More on this from Ubi founder Will Franks at www.ubiquisys.com


abhi_abhi 12/5/2012 | 4:19:47 PM
re: Broadcom vs. Qualcomm Goes Femto

In future people will read it like following:

LTE and Femto.. Qualcomm's versions of outdoor and Indoor communication

WiMax and wiFi.. Cisco's version of outdoor and Indoor communication

m__reed 12/5/2012 | 4:19:46 PM
re: Broadcom vs. Qualcomm Goes Femto

The Percello and Picochip trend is to harden as much of the design as possible, achieving the price, performance and power requirement that Keith mentions.  I really question whether TI, Freescale and Cavium can compete with their power hungry processor architectures that are not really designed for a commodotized product that Femtos need to be to achieve mass market adoption. 

I would argue that modem solutions in logic gates in an ASIC are the only way to drive down price and the TCO, which Percello have very successfully done and Picochip are starting to do.  TI already say that they are not able to compete in the commodotized business so probably some more acquisitions are to come!

Femtowatcher 12/5/2012 | 4:19:46 PM
re: Broadcom vs. Qualcomm Goes Femto


I noticed that you posted a similar comment at Seedcatalyst over the weekend.  But I still can't work out why it is that the analysts underestimate Percello's market share?  Are you saying that the femto market isn't all it's cracked up to be, so Percello's got more of it?  Or are they under-reporting their actual volumes?


Rupert_Baines 12/5/2012 | 4:19:39 PM
re: Broadcom vs. Qualcomm Goes Femto

A couple of points in comments here I wanted to join in on....

@m__reed - I would totally agree with you about the need to harden and have fully optimized SoC for mass-market. But both picoChip and Percello have already done that, with fully optimized SoCs released & shipping.  It is interesting to ask what plans are for TI, Freescale & Cavium?

@abhi_abhi: your split of Cisco = WiFi, WiMAX and Qualcomm = cellular & femto was probably true in 2007, but not now. Cisco is a major supplier of femtocells - indeed a lot of the commentary around the Informa 3Q market status pointed out that Cisco is now the #1 supplier of basestations in North America !

@KCDay (and @femtowatcher).  I am puzzled by those market share estimates: they don't seem to align with any of the reports I have seen. I would be interested to see more details or explanation. Most of the analysts seem to agree on this year being about 1.2 - 1.6M with ATT & Vodafone being significantly the largest, and Sprint, Verizon & Softbank being next.

I guess we will see: Dell'Oro, iSuppli are firms who specialise in just this, and both will be publishing reports on volume & share based on 'actuals'.

Sign In