Will Broadcom buy change the femto market?
10/28/2010 1:10 AM EDT
Implications on its competitors
How bad can Broadcom’s Percello acquisition be for
competing femtocell chip vendors? The industry knows too well what
happens to silicon startups when their core function gets sucked into a
Rupert Baines, vice president of marketing at picoChip (Bath, the U.K.),
a leader among femtocell chip suppliers, is keeping his cool. He called
the Broadcom/Percello deal “validation” of the femtocell market, while
adding: “Percello has found a good home, and we welcome Broadcom into
the femtocell market.”
However, Baines, who acknowledged the emerging market for
femtocell-integrated boxes, described the integration “a segment of the
market: not all of it.”
The femtocell integration market won’t be a slam dunk, because “most
people get cellular sevice from someone other than their broadband
supplier,” said Baines. That would make it “impossible to integrate,” he
Further, even if the systems are integrated in one box, that does not
mean that the chips will be integrated, said Baines. “We have customers
making integrated gateways today. They get the femto chip from us, and
the broadband chip from someone else.”
He said, “Part of the problem is what do you integrate?” In addition to
three femto flavors (HSPA, cdma2000 & soon LTE), there are at least
four backhaul options (ADSL, VDSL, cable, fibre, etc). Baines said,
“It gets expensive to make 12+ different die, or to ‘tax’ the people who
don't want one modem by including it anyway.”
Peter Jarich, research director at Current Analysis, agreed with Baines.
“First, it’s not clear how quickly the market will go to an integrated
gateway model. Secondly, it will still be possible to integrate other
femtocell silicon into gateways…and vendors may want to do this where
they think a standalone femtocell silicon package offers advantages or
Forward Concepts’ Strauss, however, suspected that “PicoChip is well
aware that they need additional chip solutions to broaden their market,
and it is pursuing markets other than just femtocells.”
There are also a lot of options available for PicoChip. Joseph Byrne,
senior analyst at the Linley Group, noted, “Picochip isn't the only one
at the dance in this situation.” He said that there are also suppliers
of DSL chips, such as Ikanos and Lantiq, suppliers of PON chips, such as
Broadlight and PMC, suppliers of Wi-Fi chips (Atheros, Ralink, and
Realtek), and others that would benefit from working with a company like
Picochip with femtocell technology.
Broadcom’s Fischer was cautious in outlining the company’s plan for
“silicon-level integration” for femtocells. While noting that “it all
depends on how the market moves along,” he said, “It is feasible
sometime in the next two years.” But first, Broadcom has to wait and see
which products sell, and which platform is best to integrate
femtocells, he explained.
Broadcom stressed that the company is not a total novice to femtocells.
Broadcom quietly bought Radioframe’s femtocell assets almost a year ago.
While Broadcom was often rumored as Radioframe’s mystery femto buyer,
the company never confirmed it until now. Fischer said, “We have built
our knowledge base on femtocells through a team of engineers we acquired
from Radioframe; then, Percello came on the radar.”
Industry analysts are split on their assessments of Percello. Current
Analysis’ Jarich called Percello “not as strong as PicoChip in terms of
momentum.” But he added that Percello has stronger traction than
Qualcomm, especially given its work with Ubiquisys.”
Forward Concepts’ Strauss recalled the time when Ubiquisys’ CTO told him
that the Percello chip proved to be measurably more powerful than
competitors at the time they selected Percello.
The Linley Group’s Byrne believes Percello was early with femtocell
chips supporting 8 and 16 users, which would be used in systems for
offices. But “Picochip has since met and exceeded that scale. The bulk
of the market is in smaller (e.g., 4-user) femtocells, where Picochip is
strong,” Byrne added.
Regardless who has more momentum, Broadcom chose Percello, “because we
liked Percello’s rifle-shot approach,” said Broadcom’s Fischer.
Percello’s femtocell is focused on UMTS. The company has no solutions
for LTE yet. “It is low power, cost effective and easy to integrate in
an SoC,” he explained.
By “easy to integrate,” Fischer means that the hardwired approach taken
by Percello is a good fit for Broadcom’s future SoCs. Unlike the DSPs
used by TI or PicoChip, this will make Percello’s solution infinitely
easier to integrate, he explained.
Broadcom, which has had no serious experience in the base-station market, sees femtocell as a clear opportunity to get its
foot in the new market’s door. While big boys like Texas Instruments may
have every reason to defend their big share in the base-station market
by downplaying the significance of femtocells, Fischer said, “We are not
threatened by femtocell. Rather, we see femtocell a real disruptive