x
Optical/IP

Juniper Cruises Through Q1

Juniper Networks Inc. (NYSE: JNPR) delivered as expected in its first quarter, although some questions still linger as the company enters what's expected to be a growth year.

It's an interesting time for the company, because Juniper is finally reporting full earnings statements again after dealing with some issues around stock options back-dating. Meanwhile, another handful of top executives have left, reopening questions about the executive turnover Juniper has seen in the past few years. (See Juniper Catches Up, Writes Down and Will Juniper Be Burned by the Churn?)

For its first quarter, ended March 31, Juniper reported net income of $66.6 million, or 11 cents per share, on revenues of $626.9 million, compared with net income of $71 million, or 12 cents per share, on revenues of $595.8 million the previous quarter. (See Juniper Reports Q1.)

For its first quarter a year ago, Juniper reported net income of $75.8 million, or 13 cents per share, on revenues of $566.7 million.

Juniper's non-GAAP net income of 19 cents per share matched analyst expectations as tallied by Reuters Research .

On a conference call with analysts today, Kriens said Juniper shipped more than 200 units of its core routers, the T-series and TX, during the first quarter. Juniper also got its first smattering of revenues from the MX960, an Ethernet platform that covers a hole in the company's product offerings, particularly when it comes to IPTV and similar services. (See Juniper Antes Up on Ethernet (Finally).)

"In aggregate, the market that we participated in grew faster than we did last year," Kriens said. "We have the additional opportunity to participate, with our MX products, in a segment we were not nearly as active in, in 2006."

Analysts on the call pointed out some other touchy spots though. Nikos Theodosopoulos of UBS Investment Bank noted Juniper is losing money on its Service Layer Technologies (SLT) business, which consists mainly of the NetScreen acquisition and is a key part of Juniper's ambitions in the enterprise. For calendar 2006, the SLT group reported operating losses of $12.8 million on revenues of $479.9 million.

Kriens noted that Juniper sells routers to the enterprise, too, so a bad number from SLT isn't an indictment of the entire enterprise business. Still, he admitted that "there is a gap in the contribution from the enterprise business relative to that of the service provider business." Kriens hinted that Juniper expects the SLT group to get back to profitability this year.

The first quarter saw Juniper finally cash in on $50 million in revenues from Verizon Communications Inc. (NYSE: VZ). Those revenues were deferred a year ago –- apparently for esoteric bookkeeping reasons, as the products had already been received on Verizon's end. That amount was originally supposed to be $25 million to $35 million, but it grew to $40 million with Juniper's fourth-quarter earnings, and it's now apparently up to $50 million. (See Juniper Defends Core Business in Q1 and Juniper Talks Up 2007.)

That $50 million will make for some screwy numbers. Verizon represented more than 15 percent of Juniper's revenues in the first quarter, something that's not likely to happen again. And Kriens noted Juniper will probably get a market-share boost from analyst firms that track vendor revenues.

Juniper's forecasts for its second quarter -– revenues of $640 million to $650 million, and earnings of 20 cents a share –- were on par with Wall Street expectations.

Juniper stock was down $0.95 (4.5%) at $20.16 in after-hours trading on Monday.

— Craig Matsumoto, West Coast Editor, Light Reading

Page 1 / 5   >   >>
twill009 12/5/2012 | 3:09:35 PM
re: Juniper Cruises Through Q1 It's been a long time since somebody said Avici was "on target" ;-)

Wasn't the 12000 code name originally the "Big Fungible Router"?

In my book, a commodity is something where prices and margins go up and down based on supply/demand balances/imbalances. I don't see much of that happening for Cisco and Juniper. Of course, maybe what we are seeing is tacit collusion in a commodity market controlled by a duopoly.

Can i say that?

If i count Alcatel or Huawei in the mix, does that make it a tripoli?

rjmcmahon 12/5/2012 | 3:09:37 PM
re: Juniper Cruises Through Q1 The problem for the premium router manufacturers is that SPs are looking for (and maybe finding) ways of doing the same things with fewer premium routers.

So it is not commoditization, but vulnerability to substitution that is the issue here.

Really good buggy whips never became commodities. :-)


I'd agree.

I'm reading Steinbeck's Grapes of Wrath where the Joad family leaves the Oklahoma dustbowl during the great depression to find work in the orchards of CA. They pack their family in a jalopy and when these poor farmers see all the greenery and abundance of the CA crops they're sure they'll be able to get a job and feed their families. Unfortunately this isn't the case. If they even try to take a single piece of fruit from this abundance to feed their starving children they'll be shot by hired hands of the owners and their kids will surely starve.

Startups (and VCs) chasing the service provider market reminds me of the Joads. Seeing all that cash flow that VZ and AT&T have locked up (and protected by the government), thinking they can get a piece of the action. They're pawns to these behemoths at best (such as Vonage who is of no use now that remonopolization without economic regulation has been set in motion.) Hopefully most don't have their kids bellies dependent upon their charity as these companies have little interest in any so-called "innovation."

So we'd better get used to buggy whips for awhile. That's all that we're going to get right across our backs and the backs of our children and grandchildren. Maybe the pain will take away the hunger, God knows progress won't ;-)
desiEngineer 12/5/2012 | 3:09:38 PM
re: Juniper Cruises Through Q1 mrzippy: In this case the "super control" plane is the BGP RRs, and a simpler, control plane just consisting of a single IGP instance distributing loopbacks and MPLS LDP protocols.

But BGP runs on the endpoints also as RR client. BGP RR only serves as a proxy. Of course, you can make that proxy more intelligent, but that's what it is. The BGP RR clients still have to do route resolution, still have to deal with multiple routes, etc, right?

What am I missing?

-desi
mr zippy 12/5/2012 | 3:09:39 PM
re: Juniper Cruises Through Q1 This may be the dream, but there is no sign of this coming to fruition at the moment. Fore the time being there are too many practical advantages to having the control plane and the forwarding plane tightly coupled (and strictly separated.)

Couldn't Materialgirl's description be a broader way of describing the BGP/MPLS model i.e. MPLS core only running an IGP, with the only purpose of the IGP to discover and maintain single paths to MPLS devices i.e. loopbacks or LSR identifiers, and then BGP Route Reflectors, that aren't in the traffic forwarding path, pushing routes and label to route mappings info to the MPLS LERs ?

In this case the "super control" plane is the BGP RRs, and a simpler, control plane just consisting of a single IGP instance distributing loopbacks and MPLS LDP protocols.
andropat 12/5/2012 | 3:09:39 PM
re: Juniper Cruises Through Q1 vferrari,

well said.

Pat
chook0 12/5/2012 | 3:09:39 PM
re: Juniper Cruises Through Q1 Materialgirl writes:
---------------------------
Dear Beaowulf888:
That is my point. I do not think GOOG buys L3 routers, as much as Force 10 switches. AKAM is Layer 4 or above. These guys are growing much faster than router sales would imply. Either router capacity is getting soaked up big-time, or something else is taking the load. The numbers just do not add up.
------------------------------------------

GOOG buys a lot of "premium routers" as well as L3 switches of the Force10 type.

--------------------------------
I think AVCI is on target, that the control plane is leaving the data plane, with forwarding pushed down while a super control plane is forming, splitting the router in two.
-------------------------------------

This may be the dream, but there is no sign of this coming to fruition at the moment. Fore the time being there are too many practical advantages to having the control plane and the forwarding plane tightly coupled (and strictly separated.)

--chook
chook0 12/5/2012 | 3:09:39 PM
re: Juniper Cruises Through Q1 MaterialGirl said:

Meanwhile, service providers always hated the Internet and Layer 3 forwarding. As they move to packet infrastructure, it seems like they are re-creating SONET with Ethernet at layer 2-plus instead. They seem to prefer manual configuration of point-to-point links over over automated any-to-any forwarding. This also serves to weaken the any-to-any feature of the Internet, while increasing their hold on customers.

Bottom line, the routing market is weak overall. Service providers do not like them. They are looking at layer-two solutions. Perhaps the routing market itself is rolling over, and the Internet with it. Comparing IOS-XR and JUNOS could just be arranging deck chairs on the Titanic.
--------------------------------

It's a good point, but I wouldn't go quite that far. What service providers are telling us is that they don't see putting L3 routers directly over lambdas as the right paradigm. They want something cheap and deterministic between the L3 service engines and the fibre.

In the current model, that cheap and deterministic thing is TDM (SDH and SONET) and in one view of the world its replacement is PBT.

To me PBT represents a rejection of MPLS as a replacement for SDH, not an overall rejection of MPLS and IP.

This is the second attempt at replacing SDH/SONET multiplexing (but not necessarily framing) with a packet-switched model. ATM (and its competitor, STM) were initially conceived as replacements for TDM.

One could argue that the headlong rush to turn ATM into a LAN standard and all the concomitant complexity that added was a significant factor in the demise of ATM. Nortel's Passport stuck to the TDM replacement model which could be part of the reason why it "won" the ATM demolition derby.

Probably some good hints for the devlopers of PBT in there.

--chook
chook0 12/5/2012 | 3:09:39 PM
re: Juniper Cruises Through Q1 ---------------------
Ok, trying to keep things pretty, what would be an indicator that routers are becoming (or have become) commodities? JNPR's gross margins seem strong and stable yet routers are widely available I think. Wouldn't commodity as define by Merriam-Webster suggest otherwise?
---------------------

The type of routers that JNPR makes are a long way from being commodities. Basically nobody except Cisco and Juniper (and perhaps Huawei) can make core routers. What's more, The CRS is still half-baked and maybe 2 years from being fully baked and the Huawei offering is way behind in quality.

The problem for the premium router manufacturers is that SPs are looking for (and maybe finding) ways of doing the same things with fewer premium routers.

So it is not commoditization, but vulnerability to substitution that is the issue here.

Really good buggy whips never became commodities. :-)

--chook
rjmcmahon 12/5/2012 | 3:09:40 PM
re: Juniper Cruises Through Q1 It is not a misuse of terms. You are just not up-to-date on the use of the word commodity. In my opinion, it is better to avoid the ugly word 'fungible'.

Ok, trying to keep things pretty, what would be an indicator that routers are becoming (or have become) commodities? JNPR's gross margins seem strong and stable yet routers are widely available I think. Wouldn't commodity as define by Merriam-Webster suggest otherwise?
mr zippy 12/5/2012 | 3:09:40 PM
re: Juniper Cruises Through Q1 Routers are not going away, they are getting smarter so that they can enable networked collaborative applications and new forms of networked communication like Google and Youtube.

I think what Cisco are really fundamentally doing are trying to turn their routers as a platform to run application servers on.

If you look at the sorts of application style things they're starting to push as modules or features into their routers (e.g. "ISR", load balancing modules for 6500s etc.), they're pushing application knowledge back into the network, rather than keeping it at the edge.

In the context of platforms such as the 6500s, this is good for them, because they're effectively charging customers a high premium for the Cisco proprietary interconnects that exist inside a 6500 chassis rather than a 10Gbps IEEE interconnect if the customer were to implement those services on an external and more traditional model using external servers. The only reason I can see to get one of these internal modules is if you need the sort of bandwidth for these services the 6500 backplane can provide. That being said, Etherchannel can usually solve that problem for external devices anyway. It's probably cheaper to use external devices, and you can get more redundancy and scalability by having these services reside at the actual network edge rather than just near it.
Page 1 / 5   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE