|
While
the telecom industry goes through hard times, Intel
surprises us by actively pursuing this vertical. It
thinks it can repeat the success it enjoyed in the desktop
and enterprise server space. The game plan and strategy
remains the same. Can Intel succeed? by Brian Pereira
 |
| "We
feel that telecom operators are fast moving towards
standard building blocks,” says R. Ravichandran, |
Around
the world, particularly in the US and Europe, the Telecom
players are going through a rough patch. Wireless operators
are wondering how they will recover huge investments
in telecom infrastructure, as competition fuels further
price cuts for subscriber services. Telco giants like
Ericsson are slashing their workforce in a bid to stay
afloat. Yet Intel, which sees lot of opportunity for
standard server building blocks in verticals, has chosen
to focus first on Telecom. Its Enterprise Platform Group
will take Intel architecture to platforms for Telecom
industries.
"So far there have been proprietary architectures
in the Telecom space (non-Intel architecture), and we
know the telecom players, be it the carriers, the telecom
equipment providers, the ISPs or xSPs, would need to
extend their infrastructure—that presents significant
play for Intel Architecture. This is the first vertical
Intel has got into in terms of creating platforms for
specific industries," says R. Ravichandran, Intel's
Asia Pacific Regional Marketing Manager, Enterprise
Platform Group.
OPPORTUNITY
Despite the slump in telecom markets, Intel remains
convinced that there is a lot of opportunity in the
telecom vertical, for a number of reasons.
For one, telecom players are increasingly moving away
from proprietary solutions and adopting horizontal,
industry-standard and non-proprietary solutions. Using
standard building blocks is a way to cut costs. Intel
has already proved itself in the desktop and enterprise
server space, as a provider of industry standard solutions.
Those offering proprietary architectures (like RISC)
have gradually adopted Intel architecture.
And Intel sees the same thing happening in telecom,
so don't be surprised to hear announcements about it
forging partnerships with Nokia, Ericsson, Motorola
etc. Traditional telecom players, who are into circuit
switching, may eventually relent and adopt Intel's standard
building blocks.
"We feel that telecom operators are fast moving
towards standard building blocks. They are moving away
from proprietary architectures because they need to
bring in new services due to competition. If they rely
only on one vendor they are not going to be successful,"
says Ravichandran. "Building blocks offer flexibility
and modularity, thereby allowing service providers to
offer new services and yet keep costs within control."
Already Intel has made a move to bring about a standard
operating system for telcos. The Intel Communications
group is considering Linux.
"Intel has taken standard Linux and identified
the requirements to make it more ruggedized. We have
submitted this to the Open Source Development lab. CGL
is another initiative in this direction," says
Ravichandran.
Intel is known for creating industry standards by developing
ecosystems. This creates demand for horizontal solutions,
thereby reducing market share for proprietary solutions.
Here's how Intel has done this in the past.
Intel developed its desktop and enterprise server platforms
on the basis of price/performance. It also offers customers
a choice of solutions, as several vendors have adopted
Intel architecture. Also, Intel has a broad ecosystem—lots
of software vendors writing applications. Now Intel
is hoping to replicate all this for telecom. For instance,
it is hoping that telecom application vendors (billing
software, VoIP software etc) will port their applications
to Intel platforms.
Another reason why Intel is pursuing telecom is growth
in traffic.
"The reason why we are going after the telecom
space is very clear. Internet growth or traffic growth
continues despite the slowdown. Yes, headcounts and
R&D budgets are reducing, but it does not imply
that the market is declining. Traffic is continuing
to grow, which means they (telecom players) need to
put in more infrastructure. Customers are demanding
new services (data and voice). And if the telecom players
continue to use proprietary solutions, they are going
to have major problems," says Ravichandran.
Intel is also optimistic about telecom as it envisions
the formation of a unified network in future. There
will be mobile networks and people will be accessing
these using PCs, PDAs, cell phones—though the airwaves
or broadband Internet connections.
CHALLENGES
But Intel is treading new ground and despite its strong
technical expertise and years of experience, it faces
a number of challenges.
For one, telecom infrastructure has traditionally been
the fiefdom of giants like Lucent Technologies, Alcatel,
Ericsson, Nortel Networks, Motorola and others. It could
take several years for Intel to grab market share from
these giants, who have a solid technical foundation
and lots of experience in this area.
Secondly, Intel will have to do a lot of re-engineering
on its products to create electronic components for
carrier-grade equipment—which is different and
more ruggedized than enterprise servers and enterprise
networking equipment.
Carrier-grade equipment must be designed to work in
extreme environments—at high altitudes or low temperatures;
it should be dustproof and corrosion proof; it should
also be able to withstand high temperatures (in case
there's a fire in the server room), and even vibrations
caused by earthquakes.
There are two industry certifications to check if the
equipment is really up to this. All Telecom equipment
must pass stringent tests and should be NEBS (Network
Equipment Building Specification) or ETSI (European
Telecom Standards Institute) compliant.
Though Intel has lots of technical expertise in semiconductor
manufacture, it is playing on a different field this
time (Telecom) and has much learning to do.
Thirdly, Intel and its OEM partners (HP and IBM) have
to contend with rival Sun Microsystems, which is a strong
player in the carrier grade server space. Sun's offering
here is the Netra server. Presently HP and IBM offer
carrier grade servers based on Intel architecture.
Intel purports to beat Sun using the price/performance
card. Intel claims that IBM and HP telco servers (based
on Intel architecture) have overtaken Sun servers on
the performance front (SPECint2000 and SPECweb99 benchmarks).
Also, Intel asserts IA telco servers cost 40 percent
less than Sun's Netra 20 servers.
But competition will increase as new players (like Nokia)
enter the field. It has taken years for Intel to establish
itself in the enterprise server space. But Intel won't
have so much time to reenact this in the telecom space.
OTHER VERTICALS
Besides Telecom, what could be the next vertical for
Intel to pursue? Intel offers solutions to numerous
industries, but these are standard commercial grade
servers and network solutions. Take Healthcare for instance.
Intel has set up a division called the Proactive Health
Research group, to look at how computing can be used
in the healthcare industry. This division will find
new platforms for products like tablet PCs and portable
video players. The division will also take existing
technology and use it to help senior citizens live more
comfortable lives.
Ravichandran says Intel is not going to do significantly
different engineering for hardware in other verticals,
at this point of time.
"Other verticals like Power, Medical etc are contended
with standard, off-the-shelf building blocks. These
verticals do not need completely different engineering
from a hardware point of view."
Brian Pereira can be reached at brianp@networkmagazineindia.com
|
|| Feedback
|| Subscribe-
