x
Optical/IP

Marconi's Switch Gets Tolly-fied

Marconi plc (Nasdaq/London: MONI) today announced results of tests of its BXR-48000 switch conducted by Tolly Group (see Marconi Passes Tolly Test).

The tests, on a single-rack version of the switch, indicate that it can handle 2 million simultaneous connections operating at wirespeed, with an overall capacity of 240 Gbit/s. A two-rack version is supposed to handle 480 Gbit/s, although this wasn't tested.

Impressive? Yes. But does this prove that the BXR-48000 is what Marconi claims it to be? That is, a "switch-router" that "unlike any other platform in its class... will concurrently and natively support connectionless IP routing, MPLS, and ATM switching."

Not really, on a couple of counts.

First, although Marconi says the tests were "independent," it paid the Tolly Group to conduct them. In other words, Marconi was in a position to say which aspects of its switch should be tested and which aspects should be discreetly overlooked. Likewise, Marconi was in a position to supress publication of any results that didn't quite come up to scratch.

Of course, this may not have happened. "Tolly's credibility is beyond reproach," says Geof Becker, a Marconi spokesman. All the same, there's no way of checking.

There's also some evidence that the Multiprotocol Label Switching (MPLS) and Internet Protocol (IP) routing capabilities of the BXR-48000 weren't exactly stretched to the limit. For a kickoff, the test report posted on Tolly's Website makes it clear that the current commercial version of the BXR-48000 is a plain and simple Asynchronous Transfer Mode (ATM) switch. The packet-over-Sonet interfaces tested by Tolly are in an "advanced stage of development," according to Tolly's test report (page 4). In other words, they're prototypes that aren't yet commercially available.

Other evidence of Marconi's reluctance to have the MPLS performance of its switch put under a public microscope comes from Light Reading's own test of multiservice switches, conducted by European Advanced Networking Test Center AG (EANTC) using test equipment from Agilent Technologies Inc. (NYSE: A).

In this case, the tests are truly independent. They are being paid for by Light Reading; Marconi simply had to ship its switch to EANTC's lab and configure it. And yet Marconi pulled out.

Why? Marconi's Becker says Marconi "had concerns" about the test plan and didn't have the resources to have its switch tested by both Tolly and Light Reading.

In fact, Marconi spent a couple of months discussing the test plan put forward by Light Reading and EANTC. The initial plan was heavily focused on MPLS, and to accommodate Marconi, the emphasis was shifted so that it was equally split between ATM and MPLS. Even so, Marconi ended up declining to participate.

The results of Light Reading's multiservice switch test are scheduled for publication in the next few weeks.

So far, Marconi has only managed to sell its BXR-48000 to one customer, the U.S. Department of Defense, which is reportedly using it as an ATM switch, not a multiservice switch.

Last week, Marconi laid off a further 255 people in the broadband routing and switching division manufacturing the BXR-48000, leaving about 800 on staff.

David Drury, vice president of technology strategy at Marconi and president of the MPLS Forum, was among those let go. Drury is thought to have played an important role in the DoD contract. "I didn't see this coming," Drury told Light Reading today, as he cleared his desk. "I'm still dealing with the shock and anger."

— Peter Heywood, Founding Editor, Light Reading
www.lightreading.com Want to know more? The big cheeses of the optical networking industry will be discussing multiservice switches at Lightspeed Europe. Check it out at Lightspeed Europe 02.

<<   <   Page 4 / 4
hyperunner 12/4/2012 | 9:22:33 PM
re: Marconi's Switch Gets Tolly-fied Ibeenframed wrote...

"Exactly the point - Marconi was singled out using the timing of the Tolley group report as an excuse for the thrashing."


But that's the whole point. Let's say LR had 3 vendors lined up for the switch test. Two of them pull out and one stays in. One of the vendors who pulled out (ie. Marconi) then pays Tolly to test their switch instead of using LR's free test for which they've been allowed to input suggestions on the test schedule.

The other vendor who pulled out stays quiet about it.

IMHO, LR are probably justified in assuming that Marconi has something to hide, no?

hR.
<<   <   Page 4 / 4
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE