Kathleen Nemeth Kevin R. Johnson - Chief Executive Officer, Director, Member of Mergers & Aquisitions Committee, Member of Stock Committee, Member of Offering Committee and Member of Stock Repurchase Committee Robert L. Muglia - Executive Vice President and General Manager of Software Solutions Division Robyn M. Denholm - Chief Financial Officer, Executive Vice President and Member of Stock Committee Rami Rahim - Executive Vice President and General Manager of Platform Systems Division

Ehud A. Gelblum - Morgan Stanley, Research Division Roderick B. Hall - JP Morgan Chase & Co, Research Division Eric A. Ghernati - BofA Merrill Lynch, Research Division William H. Choi - Janney Montgomery Scott LLC, Research Division Simona Jankowski - Goldman Sachs Group Inc., Research Division Kevin J. Dennean - Citigroup Inc, Research Division Ittai Kidron - Oppenheimer & Co. Inc., Research Division Mark Sue - RBC Capital Markets, LLC, Research Division

Greetings, and welcome to the Juniper Networks Second Quarter 2013 Financial Results Conference Call. [Operator Instructions] As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Ms. Kathleen Nemeth, Vice President of Investor Relations. Thank you. Ms. Nemeth, you may begin.

Thank you, operator. Good afternoon, and thank you, everyone, for joining us today. Here on the call are Kevin Johnson, Chief Executive Officer; Robyn Denholm, Chief Financial Officer; and Bob Muglia, Executive Vice President, Software Solutions Division. In addition, Rami Rahim, Executive Vice President, Platform Systems Division, is on the call and will be available for Q&A. Please remember when listening to today's call that statements concerning Juniper's business outlook, economic and market outlook, strategy, future financial operating results and overall future prospects are forward-looking statements that involve a number of risks and uncertainties. Actual results could differ materially from those anticipated in these forward-looking statements as a result of certain factors, including economic conditions generally or within the networking industry; changes in overall technology spending and spending by communication service providers and major customers; the network capacity requirements of service providers; the timing of orders and shipments; manufacturing and supply chain constraints; variation of the mix of products sold, customer perception and acceptance of our products; rapid technological and market change; litigation and other factors listed in our most recent 10-Q filed with the SEC. All statements made during this call are made only as of today. Juniper undertakes no obligation to update the information in this conference call in the event facts or circumstances change after the date of this call. In discussing the financial results, Robyn will first present results on a GAAP basis, and for purposes of today's discussion, we will also review non-GAAP results. For important commentary on why the management team considers non-GAAP information a useful view of the company's financial results, please consult the press release furnished with our 8-K filed with the SEC today. For the detailed reconciliation between GAAP and non-GAAP results, please see the Investor Relations section of our website. On today's call, Robyn will also be providing forward-looking guidance. As a reminder, guidance is provided on a non-GAAP basis other than that with respect to revenue and share count. All guidance is forward-looking, and actual results may vary for the reasons I noted earlier. GAAP guidance measures are not available on a forward-looking basis due to the high variability and low visibility with respect to certain charges, which are excluded from non-GAAP guidance estimates. Please note that today's call is scheduled to last for 1 hour. [Operator Instructions] With that, I'll turn the call over to Kevin. Kevin R. Johnson: Thanks, Kathleen, and welcome to all of you joining us on today's conference call. Before we review our earnings, I'd like to take a minute to comment on my decision, which was announced today, to retire as CEO of Juniper once a successor is named and an orderly transition is accomplished. After 32 years in a line operating role, with the past 5 years here at Juniper, my family and I have decided it was time to take a break. The company is healthy and growing, we have a strong management team and world-class talent in the domain of networking. The board has formed a search committee, and we are working closely together to ensure we manage the process in a thoughtful way. I'm committed to working closely with the board and my successor to ensure a smooth transition for all of our stakeholders. Until then, I continue to serve in my current capacity. I will also remain focused on fulfilling my board commitments through this process. It is an absolute privilege to serve as Juniper's CEO, and I've enjoyed the past 5 years. Together, we have grown our business and strengthened our position in the networking industry as evidenced by this quarter's results. The business is healthy and momentum is building. So let's spend some time to talk about the quarterly results. We delivered strong second quarter results, reflecting good execution on our strategy, as well as the strength in our product portfolio. I'm pleased with the performance of this quarter in what remains a dynamic global economic environment. To put these results into context, you may recall that for the past 3 quarters, we have been talking about demand improving in service provider. We saw good strength in service provider this quarter, both in terms of revenue and certainly in terms of the bookings environment. I'll speak more about the service provider segments, carrier, cable and content in a moment. In enterprise, while there continues to be areas of weakness, we are seeing some early signs of stabilization and improving demand in the U.S. and Europe. In service provider, we focused on 3 customer segments: carrier, cable and content, also referred to as Web 2.0 providers. Carrier includes both wireline and wireless. In the second quarter, we experienced good overall demand in wireline, while demand also was strong in wireless in the Americas. In EMEA, we are seeing emerging signs of growth in wireless, we performed well in cable and we are seeing great traction with major cable providers. In content, our good traction in core routing and switching are translating into improved performance as CapEx strengthens in this segment. For the routing market, we continue to see very strong demand in the Edge. Core networks seem to be running hotter, which is a positive indicator as we expect investment in core routing begin to pick up in the coming quarters. Overall, the trend we identified 3 quarters ago of a routing CapEx cycle appears to be picking up. On the enterprise side, our switching and data center business performed well in the second quarter. Highlights included strong demand from our newly introduced EX9200, as well as our QFabric family of products, which had a record quarter. Customer response to the EX9200 has been good, as it satisfies our customers need for programmability in core switches that power both campus and data centers. This also aligns with our SDN strategy. In security, we're seeing positive activity in bookings. Bob will provide more detail on our strategy and progress in a few minutes. As a company, we continue to drive innovation and differentiation through our systems and our software. We're focused on delivering world-class scale, performance and reliability in our systems, which is the heritage of our company. At the same time, we're complementing the value delivered through our systems with a software strategy that enables our customers to be more agile. This is delivered with our network function virtualization capability in a way that helps customers reduce cycle times and be more efficient, and we expect to enhance this capability with future product introductions that are part of our SDN strategy. We believe the combination of scale and performance in our systems, plus the agility and efficiency in our software, provides unique value that is resonating with our customers. The second quarter was a strong one for revenue, bookings and backlog. In addition, we remain focused on prudent cost management and effective capital allocation, which Robyn will discuss later on the call. To sum up, our 2013 results show that we are successfully executing on the 5 operating principles we established at the beginning of the year. We expect macroeconomic environment to remain uncertain, we expect modest growth in the markets we serve, take share in routing and switching and stabilize share in enterprise security, expand 2013 operating margins over 2012 and continue to generate solid cash flows and prudently allocate capital. The strength that we're seeing in the service provider market and the stabilization we're seeing in the enterprise market reinforce our confidence that we are in the right markets with a great product portfolio. We continue to feel very good about the health of the business, and we will continue to stay focused on executing with agility and delivering value for all our stakeholders. Now I'll turn it over to Bob to discuss the latest developments in security and SDN. Bob? Robert L. Muglia: Thanks, Kevin. Today, I'd like to update you on our progress in the security business, as well as with SDN and our overall software strategy. As we've previously outlined for security, we expect our recovery to occur over multiple quarters. With that background, Q2 is a tough quarter for our security business, with revenue of $126 million, a decline of 20% year-over-year and 8% quarter-over-quarter. However, the actions we are taking are showing some positive signs and results. Our security bookings grew 6% over Q1 2013, and we exited Q2 with a strong backlog in our security business. We are also seeing encouraging growth in our security sales pipeline. Providing color on this, as previously outlined, you can divide the network security business into 3 distinct customer value propositions: securing service provider customers, securing corporate end users within the campus and branch and securing the data center from Internet hackers. The primary usage of network security for service provider customers is enabling high-scale firewalls for mobile users. Juniper has benefited from the growth of LTE networks as the scalability and reliability of our SRX product is ideally suited to this application. During 2013, we have seen this business slow in North America as the large service providers focused their capital spend on other areas, including Edge routing. Moving forward, we expect to see growth opportunities in EMEA, South America and to a lesser extent, in Asia, as these carriers roll out LTE. We have a leadership position in service provider security, and we intend to maintain that as more carriers around the world modernize their mobile networks with high-capacity LTE service. For Campus and Branch, we continue to see the impact of the product transition from our older ScreenOS product line to the current SRX products. Our ScreenOS products continued to decline, but we have seen good year-over-year growth in our branch SRX products. That said, this is the area where our products fell behind in the transition from ScreenOS to the SRX, and we expect recovery in Campus and Branch security will occur over multiple quarters. While we still have work to do, this transition is well underway. In Q2, ScreenOS products accounted for approximately 20% of overall security revenue. For the data center, in Q2, we shipped the highly innovative Spotlight service that works in conjunction with Junos WebApp Secure. Spotlight is the only cloud platform in the world that shares fingerprints of individual attackers' devices. This intelligent service protects our customers by blocking attackers before they have a chance to do damage. The combination of Web AppSecure and Spotlight provide a foundation for differentiation of Juniper's overall data center security products. This strategy is working for us, generating considerable interest from major enterprise customers, including the Chicago Mercantile Exchange and Interactive Data, who have chosen Juniper to secure their data center environments. In Q2, we also shipped Junos DDoS Secure, a unique application-centric service that provides protection against Distributed Denial of Service attacks. DDoS Secure was chosen by First American Bank to secure and protect their financial systems, and this win further strengthens Juniper's relationship with this important customer. We are seeing strong interest in using our security products to protect enterprise data center. And to further capitalize on this, in Q2 we took steps to restructure and grow our enterprise security sales force. We expect those steps, together with the introduction of new and improved products, to bring stability to our security business in the second half of 2013 and enable a return to growth in 2014. Switching to SDN and our broader software business, we have established a strong industry thought leadership position with our SDN strategy and products. That strategy is resonating with customers, industry analysts and the press as demonstrated by the positive feedback and reviews we perceived, as well as our overall share of voice within the press, which far outstrips what might be expected from a company of Juniper's size. In May, we announced that we had entered beta with our Contrail SDN controller and we pulled in the ship date of this product from 2014 to second half 2013. Pulling in ship date of a new product is an unusual event in this industry, and the progress we've since made with our beta customers further increases our confidence that we will meet our commitment. In terms of those beta customers, they are relatively evenly divided between enterprise and service providers, and include a strong representation of customers in all 3 global geographies. We see several value propositions emerging for SDN across both the enterprise and service providers. Within the data center, as customers implement private and public cloud infrastructures, they must virtualize their network to enable elastic scaling of resources. Enterprises want to build hybrid clouds that connect their private cloud to commercially providing public clouds for disaster recovery and on-demand availability of resources. Service providers seek to improve service creation agility within their Edge network by implementing network function virtualization. All of these capabilities are enabled by our Contrail controller, and with the beta trials that are underway, we have established strong mindshare and experience in this important emerging area. Interestingly, many of these trials bring other Juniper services, such as Firefly, our virtual firewall. While we don't expect to see material SDN revenue in 2013, we are well positioned to benefit from the growth in this area in 2014 and beyond. For our broader software business, we achieved record Q2 bookings, which is reflected in backlog, and this will be recognized over multiple quarters. We expect to see the trend of ratable revenue recognition continue as we move our software licensing to Juniper Software Advantage; the networking industry's only comprehensive software licensing model. Earlier this month, we moved the majority of our MX application software to Juniper Software Advantage, and we anticipate transitioning more of our security products in Q4 of this year. All of these steps, rebuilding our security business, establishing leadership in SDN strategy and products and moving to our new Juniper Software Advantage licensing model are meant to build a foundation for the future. It will take time, but we are on our way. Now let me pass this on to Robyn. Robyn M. Denholm: Thank you, Bob, and good afternoon, everyone. I'm pleased to report that our second quarter results reflect strong revenue and even stronger earnings growth. Revenue exceeded our expectations due to 2 main factors that were roughly of equal size. The first being increased service provider spending momentum and improved enterprise demand, and the second factor was the earlier-than-anticipated recognition of some U.S. federal revenue. Operating income and the EPS both improved significantly, reflecting the leverage in our financial model. Looking at our demand metrics, we had a good bookings quarter, with strength across our customer base, resulting in book-to-bill greater than 1. Product backlog, which you've already seen at healthy levels, increased both sequentially and year-over-year. Product deferred revenue was down modestly quarter-over-quarter. Total revenue was $1,151,000,000, up 9% sequentially and 7% year-over-year. The sequential growth was driven largely by enterprise, whilst the year-over-year growth was balanced between service provider and enterprise. No single customer accounted for more than 10% of total revenue in the quarter. For the second quarter, GAAP diluted earnings per share were $0.19. Non-GAAP diluted earnings per share were a strong $0.29, up $0.05 sequentially and $0.10 year-over-year. Now let me provide some color on revenue by region, market and business segment. Americas revenue was up 14% sequentially and 15% year-over-year. Enterprise had a very good quarter, driven by federal and financial services that resulted in double-digit sequential and year-over-year growth. Federal had a good quarter even without the benefit of the deferred revenue recognition. The year-over-year strength in the Americas was due to another quarter of double-digit service provider growth. This growth reflects the improving broad-based demand across cable, content and carriers. EMEA revenue was up 4% sequentially and 1% year-over-year, driven by enterprise strength, while service provider's revenue was flat sequentially. We did see improved demand in Northern Europe and in Germany. APAC revenue was down 1% sequentially and 7% year-over-year. We saw soft demand in enterprise across the region, and Japan continues to be weak. We did see sequential growth in service provider, primarily in China, Taiwan and South Korea. Looking more closely at the markets we address, service provider revenue was $726 million, up 2% sequentially and 7% year-over-year. USSP was strong, especially in cable, content and pockets of strength in EMEA and APAC. Enterprise revenue was $425 million, up 23% sequentially and 8% year-over-year. Sequential growth was driven by federal and financial services in the U.S. and broad-based growth in EMEA. Now let me review our revenue by segment. Platform Systems Division revenue was $916 million, up 13%, both sequentially and year-over-year. Total router product revenue of $578 million was up 12% sequentially and 14% year-over-year. The MX product line had another record revenue quarter, growing 19% sequentially and 43% year-over-year. This reflects continued strong customer acceptance of our Universal Edge offering. In addition, we recognized our first revenue on the recently released MX2020. Total switching product revenue was a record $160 million, up 22% sequentially and 15% year-over-year. In the data center, the newly released EX9200 is off to a good start and QFabric continues to generate momentum. SSD revenue of $235 million was down 6% sequentially and 11% year-over-year. Total security product revenue was $126 million, down 8% sequentially and 20% year-over-year. This area of our business remains challenged. However, as Bob mentioned, we are seeing good customer activity and response to our new initiative with some early signs of the business turning around. In the second quarter, both orders and revenue from new products increased sequentially, and we feel good about our ability to achieve our yearend quarterly run rate of $150 million. Moving on to gross margin and operating expenses, non-GAAP gross margins for the quarter were 63.7% compared to 64.6% last quarter and 63.4% a year ago. Non-GAAP product gross margins were 63.8% and non-GAAP services gross margins were 63.3%. Gross margins for the quarter would have been roughly flat sequentially, except for the negative impact of the Federal-deferred revenue recognition. We are pleased with the progress that we have made in our supply chain and our services cost reduction. Non-GAAP operating expenses decreased $3 million sequentially and $5 million year-over-year to $515 million. OpEx was at the high end of our guidance, reflecting good progress in cost reduction, partially offset by higher variable compensation and legal expenses. OpEx as a percentage of revenue was 44.8% versus 48.9% last quarter and 48.4% in Q2 of last year. Looking at our headcount, we ended the quarter with 9,513 employees, a sequential increase of 191, primarily in lower cost ranges. Non-GAAP operating margin for the quarter was 18.9% compared to 15.7% last quarter and 16% last year. Operating income increased 31% sequentially and 35% year-over-year. This reflects our strong focus on achieving our long-term model by driving operating leverage with good top line growth and OpEx management. The non-GAAP tax rate was 27.4% compared to 19.8% last quarter and 30.7% a year ago. As a reminder, last quarter, the tax rate reflected a onetime benefit of the 2012 R&D tax credit. Looking at the balance sheet, we ended the quarter with $2.8 billion of net cash and investments. For the quarter, we generated strong operating cash flow of $284 million, in line with our historical pattern. DSO was 40 days in the quarter, down from 45 days last quarter, due primarily to better order and shipment linearity. Capital expenditures were $71 million, down 1% sequentially, while depreciation and amortization was $42 million. We have now largely completed the construction of our Sunnyvale campus and also consolidated some of our labs and office space in our existing Blade facility. We repurchased 6.4 million shares for $106 million. We have $333 million remaining on our current authorization. As announced, the board has approved a new incremental authorization of $1 billion, and we expect to continue calibrating our buyback in future quarters with market conditions. Now I will review our outlook. As a reminder, these metrics are provided on a non-GAAP basis, except for revenue and share count. Our outlook for the September quarter reflects our expectations of continued good service provider demand in the U.S. and modest improvement in EMEA. We expect routing and switching to remain strong, while the security business begins to stabilize. For the third quarter, we expect revenue to range from $1,140,000,000 to $1,180,000,000. Gross margins are expected to be in the range of 64.5%, plus or minus 0.5%. Operating expenses are expected to be $525 million, plus or minus $5 million. As we have demonstrated, we are focused on driving leverage in our financial model and continue to bring our cost structure in line with our long-term operating model. As an update to our cost reduction activity, we've experienced high labor claims all year, and given the improved demand environment, variable expenses will be higher than we anticipated for the year. We also are making some incremental investments to maximize our top line growth. With that in mind, we now see a net year-over-year OpEx reduction in 2013 of approximately $50 million. We are well ahead of our $50 million cost reduction goal. At the midpoint of guidance, operating margins are expected to be roughly 19.5%. This is expected to result in non-GAAP diluted EPS of between $0.29 and $0.32 per share, assuming a flat share count and a tax rate of 28%. To summarize, we expect the demand environment to continue to improve. We are confident of our product portfolio in routing and switching and improvements in security. We are focused on executing to drive revenue growth, maintain healthy gross margins and manage OpEx to achieve our long-term model. I would like to thank our employees for their continued dedication and commitment to successfully executing our strategy. Now let's open the floor to questions.

[Operator Instructions] Our first question is from Ehud Gelblum of Morgan Stanley. Ehud A. Gelblum - Morgan Stanley, Research Division: I've got a bunch of questions for you, so hope you're ready. Question one is on the switching business. That was the strongest, I think we've ever seen, at $160 million. Can you give us a breakout on -- you said QFabric was strong. How strong was QFabric? Is it still in the order of small numbers or substantial? Kind of just give us a sense as to how big it is. And when you talk about QFabric, is it at the top-of-rack switches or are people buying the directors? And clearly, I assume it's more than minis than the original QFabric. And if you can give us a sense also as to how much of your strength in switching was from data center versus campus LAN, anything you can tell us about WiFi, just breaking out switching. And then on security, when you look at your 2 businesses in security, the carrier side and the enterprise side, Bob mentioned that the carrier SRX was not doing as well this quarter because route -- carriers are focusing more on routing. Does that mean that the 2 are kind of out of sync with each other and you can never get the 2 on the same page? And then finally on enterprise, I'll let someone else take care of that. Kevin R. Johnson: Ehud, what was your last question on -- was it on enterprise security? Ehud A. Gelblum - Morgan Stanley, Research Division: It would have just been how -- is it possible to separate that from the carrier security side and sort of maximize that for cash as opposed to -- run it for cash as opposed to for growth, what you're doing in carrier? Kevin R. Johnson: Okay, got it. Thanks for your questions. Let me sort of kick off on the first one on the switching business. I think we had a very strong quarter on switching and I think a part of that was the introduction of the EX9200, which replaced our EX8200 switch. And the difference in the EX9200 is the programmability that, that switch enables. And it is a great switch both for data centers and we see some customers even using it for campus because of the programmability. And so I think the product transition from the 8200 to the 9200 was one contributor to the strong performance in switching. And as we noted, QFabric had a record quarter. And we talk about QFabric as the QFabric family of products, which does include the top-of-rack switches, it includes the smaller version of the Interconnect as well as the larger version of the Interconnect. And this is a quarter that we had -- we certainly had continued sales and good performance of top-of-rack, but we also have very good performance of the Interconnects, both the lower end and the higher end. And Rami -- we don't break out the specific numbers, Ehud. But, Rami, do you want to comment more on the switching part of the business?

Yes, I'd be happy to. On the question of the particular segments or subsegments of switching, I can tell you, for example, with the EX9200, the early traction that we saw was a very nicely balanced between campus and data center. And it actually fits really nicely with the strategy that we've been executing towards. In fact, what we found is that many of our customers deploy our switches in scenarios where it's a common core that's used for both the data center and the campus. The software that we're introducing for our switches in the form of network director as part of our broader SDN strategy is very much about converging the management of the switching environment across these multiple areas as well. So we were pleased with the performance we've seen certainly this quarter. Robert L. Muglia: Okay, and thanks for your question. I mean, with regards to the relationship within the SP environment between routing and security, as Kevin has described in a number of times, we see different trends within service providers of areas of capital spending and those change over a period of time. And in the case of security, the area that is most strongly influenced is it follows the RAN, the build-out of the RAN of wireless carriers, as they increase the availability of spec -- and they use their spectrum and increase the ability to move on to newer networks such as LTE, that is where we see security being most closely correlated to. And what we've seen now is in North America, some of the major carriers have largely gone through their LTE build-outs. And so we see opportunities in other geographies around the world. Now the interesting thing here is just that as the capacity increases of the radio network and that is being secured, obviously, there's a need for Edge services and that's a routing thing. And as we've discussed over time, that makes the cores run hotter, and so there's opportunities there. So the correlation is really based on how the carriers are doing their capital allocation. Kevin R. Johnson: I'll just comment on the last part of your question which was if we just focus on enterprise security and didn't do carrier, would that be a way to run the business for cash, I think was the term, used in your question, Ehud. And the fact is the high-end SRX is the same high-end SRX that we used to secure these LTE networks as we used to secure large data centers. So there may be some slight differences in the features, software features that go -- are used in the high-end SRX, but fundamentally, the incremental R&D required to be in security for the LTE mobile networks is not significant -- significantly different or more from what you would do in data center. So we think it's the best return on investment to invest in that R&D and then be able to monetize that same product in both service provider and enterprise.

The next question comes from Rod Hall of JPMorgan. Roderick B. Hall - JP Morgan Chase & Co, Research Division: I just had a couple of questions for you. First of all, I wonder -- just kind of housekeeping point, Robyn. Could you give us any quantification on that deferred revenue from federal and let us know if there is anything in guidance for further unwind the deferred revenue there? And then secondly, I just wondered on the security business. I mean it feels like we still saw sequential decline in security revenues but it feels like you've got a grip on that market now a little bit better. I wonder, could you give us some ideas, do you expect that market revenues or that your revenues from security to be flat quarter-on-quarter in Q3 or maybe even up a little bit? Are we still expecting a little bit of decline there? And then lastly, you guys seem more and more optimistic about the routing cycle. I wonder if there's any more color you can give that on that. Are you seeing particular regional movement? Is it mostly in North America, do you see routing traction in Europe? Could you just give us some more on the routing cycle that you see come in the second part of the year? Robyn M. Denholm: Okay. Thank you, Rod. In terms of the deferred revenue, the way I characterize that, the upside in the quarter for Q2 was half related to earlier-than-anticipated deferred revenue recognition and half from stronger demand from enterprise and service providers. So that's how I characterize it. As you can see from the balance sheet, the deferred revenue is only modestly down quarter-over-quarter, and so you can see that we're adding to the deferred revenue pool, as well as recognizing deferred revenue in the quarter. And the deferred revenue balances are healthy, the backlog is healthy and the bookings were healthy in the quarter. So our view is that in terms of the underlying strength of the business, even if you take out the deferred revenue recognition item, it was healthy. And so on the security side, I'll let Bob answer that question. Robert L. Muglia: Yes, Rod. In terms of security and the expectations for the rest of the year, as we said we expect the security business to stabilize in the second half of this year with a return to growth next year. In terms of whether it will be flat or up or what over the next quarter, the variability here is perhaps associated with the split between the decline that we've had in the enterprise and the service provider. Q2 was just like Q1 in the sense that the decline was structured as 2/3 of service provider and 1/3 enterprise. And we have seen some good traction on the enterprise side. In the service provider, which, just the very nature of that business is fairly lumpy. So it's difficult for me to predict on a quarter-to-quarter basis exactly what we'll see, but overall, we do expect stability going forward for the rest of this year. Robyn M. Denholm: And then routing, Rami, would you like to comment?

Yes, I'd be happy to. So just to provide a little bit of color, certainly, we've been saying for the last quarter or 2 that the majority of the drivers on routing have been in the Edge. That continues to remain true throughout Q2. And you have to understand that the drivers for investment in the Edge are just more diversified than that in the core. So people invest in the Edge provide new services, to keep up with capacity, to provide connectivity to various different types of customers. So we had a very good Q2 in terms of the Edge routing. From a geo standpoint, North America is definitely sort of a strong area for us, but we are continuing to see good signs in parts of Europe, both from a standpoint of just the market in general, but also in terms of the design wins that we have gotten over the last several quarter that we are now executing towards delivering for our customers. And then just from a segment standpoint, we saw a good balance of investment in routing, both in our cable operators, our content providers, our Web 2.0 and actually also our telco -- traditional telco. So cable operators are dealing with an influx of video traffic that they need to support, telcos, with the sort of trends of convergence and delivering on new services and content providers for data center interconnect, for peering and that sort of activity. So really nice balance of business across a large segment of our customers.

The next question is from Tal Liani of Bank of America Merrill Lynch. Eric A. Ghernati - BofA Merrill Lynch, Research Division: This is Eric Ghernati for Tal. Just a quick question on the enterprise again. I guess the strength came from federal financials with federal being due to deferred revenue recognition. These 2 verticals, if I recall, were excessively weak last quarter. I guess how should we think about the sustainability in the second half? Will these 2 verticals stay strong or will they -- other verticals or geos will strengthen? And then just, if I recall correctly, AT&T and Verizon were 10% customers. You said that you had no 10% customers. I assume that affected some degree core routing business, but you voiced your confidence that core routers would rebound in the second half. Are you seeing that in your bookings? What are your customers showing you, specifically the U.S. large telcos? Kevin R. Johnson: Thanks, Tal. So the first question you had was on enterprise. And we posted an 8% growth in enterprise. Certainly, as Robyn mentioned, I think our product deferred revenue was down a net of $9 million sequentially. So certainly, there was some benefit to the U.S. enterprise business from the recognition of that deferred revenue in federal. But even net of that deferred revenue that was recognized in federal, we still saw improving -- improvements in the performance of federal. And so things seem to be stabilizing a bit more and actually getting on more of an approved footing in federal. And we expect that to continue into the second half. EMEA posted a good growth rate in enterprise as well, and the place that I think we're underperforming a bit is in Asia. And I think we've certainly seen some challenges in certain countries like China and certainly some others, but I think we have opportunity to improve our execution in Asia a bit. But I think we saw a good, strong enterprise performance in the Americas and EMEA. And certainly, I think the fact that switching was up 15% was a part of contributing to that enterprise performance. On the second question related to core routing, I commented that part of the analysis we do is we look at the total addressable market spend against core and Edge relative to the traffic growth that we're seeing on the networks. And from that, we can start to deduce how much of that was replacement equipment versus new capacity. And what we're seeing now on the Edge is it's now building out new capacity to handle what was networks that were running hotter and hotter. And we think that, that trend is going to continue for several quarters. We have seen the core of the network in our analysis has been running hotter and hotter, and we're just now starting to see the signs that investment will, in addition to being on the Edge, will start to kick in for the core. And we expect to see that start to unfold later this year. Rami, I'll let you comment further on the core routing side.

Sure, the drivers for core are really all around capacity. So customers will tend to defer that for us long as they can for they absolutely need to investment, and we've certainly seen some of that. I'll just add that, in addition to what Kevin said, I'll just add that I think that the strategy that we have been outlining, which is around the diversity of our products in addressing different types of core challenges, whether it be the T Series as a very seamless upgrade path to an existing Q Series network or the MX, especially now with the MX2020 and a denser line card or the PTX, where we've just introduced the PTX3000 -- or we've announced the PTX3000 to be shipped in the second half of the year is resonating with our customers. And as Kevin said, we are seeing early traction that this strategy is right in terms of what our customers require.

The next question is from Bill Choi of Janney Montgomery Scott. William H. Choi - Janney Montgomery Scott LLC, Research Division: First, on the clarification part here. Did I understand here, after the deferred rev recognition on fed, you're still looking for a sequential improvement into the second half, meaning that you will benefit from the potential fed budget flush specifically in the September quarter? And then secondly, I want to understand this reinvestment in OpEx for growth opportunity to, I guess, just 6 months ago, you guys were trying to cut cost; now you see some opportunities for additional growth. And I just want to understand where that's coming from. I mean, part of this improving demand certainly is that you already had the footprint, already had the relationships of those customers come back to spend. Those don't require incremental OpEx investment to capitalize. Is there a lot more opportunities that you're seeing in this current environment for new footprint growth and new projects that require investments? I'm sorry, and then how long will it take to get before you get some of the revenue momentum from the increment investment? Robyn M. Denholm: Thanks, Bill. I think that last bit didn't make the buzzer, but anyway, in terms of the first part of the question, let me just clarify. What Kevin was talking about was in the second quarter, absent the federal revenue recognition items that we talked about, if you took that out of the numbers, we would have still been up in federal in the second quarter. And we are expecting a continuation of that trend. In other words, sequential increase in terms of what we saw between Q1 and Q2 on the base demand as continuing for the second half. So, and I might add that we also saw good financial services improvements in the second quarter as well. And Kevin also mentioned EMEA enterprise was stronger in the second quarter. So I think that in terms of that underlying business trend, we see enterprise demand strengthening. In terms of the investments, what I talked about in terms of OpEx, so for the full year, we will be down year-over-year $50 million in terms of operating expenses, 2013 over 2012. We have been very focused on continuing to drive cost reductions across the board, and we are still driving those. Having said that, I have reduced the total year-over-year decline from roughly $100 million to $50 million, and there are 2 major factors in that. The first is the continuation of the legal and variable expense increases that we've seen versus our expectations at the beginning of the year. And then the second is some modest investments that we are making in the second half to capture both the near-term and the longer-term revenue opportunities that we see. And so we think that's a very prudent thing to do from a company perspective to continue that revenue momentum as we go forward. Having said that, our OpEx as a percentage of revenue has come down dramatically from last year, and we're continuing to focus on that in line with our long-term model that we put out there this time last year. And I'll remind you that our long-term model for OpEx as a percentage of revenue is somewhere between 39% and 42% of revenue. We still have some ways to go to get to 42% being the upper end of that range, but we do expect to exit this year close to that upper end of the OpEx as a percentage of revenue range.

The next question is from Simona Jankowski of Goldman Sachs. Simona Jankowski - Goldman Sachs Group Inc., Research Division: So your service provider business in aggregate was up 2% sequentially, which was about $13 million. And I think you commented that MX Edge routing was up 19% sequentially, which is probably more than $50 million. So it seems like some of the other categories within service provider were down sequentially and I just wanted to see if you can provide some detail of how that will split between core routing and SRX within the service provider vertical, and specifically if core routing declined. And since I know you've commented on networks getting hotter in the core for the last 3 quarters, just wanted to get some specificity there on either bookings or backlog or any kind of customer activity that would suggest that, that is likely to improve. Robyn M. Denholm: So I'll start the answer to that question, Simona, and then I'll hand it over to Rami to talk about some of the trends. So in terms of our SP business, it was up 2% sequentially, nicely up year-over-year. In terms of the trends there, yes, MX was up. As you know, we have some older Edge router products that have been declining and they continued to decline in the second quarter. The core revenue was down sequentially, but the bookings was strong in terms of core. And as Bob mentioned in the security area, that was -- the SRX was also down sequentially in terms of the security product into SP. And so in terms of the overall trends, Rami can focus on that as well.

Yes, not much more to add. I'll just say that on the Edge, certainly the MX had been very strong and the -- a component of the universal Edge strategy that we are executing towards actually involve that other Edge product -- or some of the older products that we were working on in the past will decline. So that's completely anticipated. As far as Edge versus core, I think we've outlined what we're seeing in terms of the drivers in the industry, and these are things that play out over multiple quarters. I don't anticipate any major changes any time soon. That said, we did say that we're starting to see early signs that the core investment cycle might be starting and bookings is one metric that we're using as a reflection of that. Simona Jankowski - Goldman Sachs Group Inc., Research Division: And Rami, just as a follow-up to that, are there opportunities when you think you might be selling a PTX or an MX or MX2020 in place of what would have been a T4000 opportunity? And if that's the case, is that $0.50 or $0.80 on the dollar? And then along the same vein, with the new routers and core from Cisco and Alcatel, what is your early take on how that is impacting the marketplace, either from a market share or a pricing perspective?

Yes, to your first question, I mean, the way I look at this is we have a tool bag full of tools that are disposable, that we can leverage when we have the conversation and we go after an opportunity with our customers. So any core opportunity starts with a conversation between ourselves and our customers, trying to understand their business challenges, what they're trying to achieve. And then we can go, based on that understanding, reach into our bag to determine whether we're going to be satisfying their requirements with the T Series or an MX or a PTX. And that is a very positive thing because, quite frankly, our competitors don't necessarily have all of those options available for them. So that diversity of offering that we have, the strategy that we're executing towards I think is really a powerful one. The second question that you asked is around... Simona Jankowski - Goldman Sachs Group Inc., Research Division: Alcatel and Cisco.

Yes, competitive situation. I'll say sort of what I said in the last earnings call, which is that I'm not going to underestimate any of my competitors, Cisco or Alcatel. Certainly, Alcatel is trying to get into the core space. But that said, I have great confidence in the strategy that we're executing towards. I think the success that we've seen with the early products, for example, with the PTX as a very dense next-generation transport device that really our competitors do not have. The new multi-chassis offering that we just introduced for the T Series -- T4000, I should say, the early traction there, again, give me very good confidence. So we won't underestimate our competitors, but I think that the strategy that we're executing towards is a winning one.

The next question is from Kevin Dennean of Citi. Kevin J. Dennean - Citigroup Inc, Research Division: Just going back to expectations for core routing, it's been asked a few different ways, you're obviously sounding more confident on the outlook. Rami or Kevin, I'm just wondering if you could talk about the expected duration of the core routing cycle. Is there any reason to believe that this is more than just a couple of quarter pickup, or is this going to be something more durable along the lines of a multi-year cycle? Kevin R. Johnson: Yes, I guess I'll start, and then I'll let Rami complement with additional comments. In our analysis, it's been several quarters that core and Edge routing had been running hotter, and that's partly why the addressable market for routing declined 3 percentage points in 2012. It wasn't just a 1 quarter networks were running hotter, it was several quarters. And that really started mid-2011. So there was probably a 6-quarter period where, in our observation, core and Edge networks were running hotter. We started to see the pickup on Edge network investment in Q3 of last year. We saw the signs of it, and we talked about that and that led to the results that we delivered in Q4 and what we're seeing today. And we think that as Rami pointed out, the Edge investments go beyond just incremental capacity. It goes to enable new services and other things that our customers are trying to do. And so we believe that Edge cycle is going to continue for quite some time. And what we're now seeing is early signs that the core investment cycle is going to pick up, and that's got to pick up just to address capacity issues. As you build out more Edge, it's following more traffic in the core. And certainly things where a customer will build a metro packet core, basically a core within a metropolitan area to keep traffic within that area, we see projects in those areas that will, in some ways, they'd put in a core router in that metro -- in that metropolitan area. So the net is, I think the core -- the signs we're seeing will lead to increased CapEx on the core. You cannot exactly predict when it's going to occur, but the early signs based on what we're seeing in bookings in our discussions with customers on projects tell us that it's coming soon. And I would suspect it too will be multi-quarter, multi-year process as they build out for what would be 1.5 years to 2 years of running those networks hotter. It's going to take some capacity to catch up with it. Rami, I don't know if you have anything else to add or...

Not much more. It's very difficult to predict the timing and the duration of the timing. I think the thing that's in our control is to just make sure that we are best equipped to capture the opportunity when it comes about. That's exactly what I'm focused on, that's what my team is focused on. And I think things that we are going to be prepared, whether it's a solution to converge packet off, because I think we have a fantastic one, or a migration to it in a T Series capacity, I think we've got a great story there as well. So I think that's the thing that we're focused on right now. Kevin J. Dennean - Citigroup Inc, Research Division: But Rami, just a follow-up, is there anything that you see structurally different in network architectures that would point towards a shorter cycle in core routing versus what we've seen historically?

One can always have an architectural discussion around capacity in the metro versus capacity in the core. And things could evolve over time. But I'd say, at this point, no. There's nothing that points to a radical change in the architecture that is going to change the balance of metro versus core. It might be the case that you'll see some of that happening in some customers and others, but there's no sort of obvious trend that I'd say at this point.

The next question is from Ittai Kidron of Oppenheimer. Ittai Kidron - Oppenheimer & Co. Inc., Research Division: I wanted to ask about going back to this CapEx routing cycle. In the past, you used to estimate the roughly the seasonal spending pattern as 45% in the first half of the year and 55% in the second. Is this something you subscribe to this year? And if this routing cycle does happen, is anything in that pattern going to be different from your standpoint? Kevin R. Johnson: Well, I think the routing cycle is happening, and it's more focused on Edge and the core will kick in. And I think in 2013, we expect sort of typical seasonal patterns on routing CapEx. So we expect to see an uptick in the second half. Ittai Kidron - Oppenheimer & Co. Inc., Research Division: Okay. And lastly, regarding the switching business, good results on numbers there. I guess the question is on the EX9200, when you look at the 82 and the 92 together, was there growth in that business or most of that business was really a replacement factor of the 82? And second, regarding your QFabric business, it's been, although the mini is relatively new, the portfolio itself is a little bit longer in the tooth here and with Cisco introducing a lot of new products just a month ago, when do -- when would be the right time to think about an upgrade of your high-end data center switching portfolio?

Yes, so we're pleased with our switching performance in Q2 overall. Certainly, as we introduce a new product like the EX9200, the transition is a very important thing for us to think about as we get our customer sort of thinking about what's the right product for them. We're pleased with the way that, that transition had been rolling out, especially in the Q2 timeframe. And then the 9200 in and of itself is a very programmable device that's really ready for at the end. That message is really resonating with our customers. So I'm optimistic about the growth potential of that platform. And then on the QFabric, the QFabric remains -- I mean there are many that are talking about fabrics today, but the QFabric remains the single true converged fabric architecture for the data center, that provides us with thought leadership, it opens up doors for us all the time with our customers and again, the introduction of some of the smaller versions of the QFabric and the new features have provided us with some nice momentum. And last I'll say about -- when we think about the upgrade and we're always, at Juniper, thinking about next-generation upgrades and roadmaps in technology and so it's all the time.

The next question is from Mark Sue of RBC Capital Markets. Mark Sue - RBC Capital Markets, LLC, Research Division: Kevin, I hope we didn't wear you out over the past 5 years. Kevin R. Johnson: No, no, not at all, Mark. Mark Sue - RBC Capital Markets, LLC, Research Division: Okay. So my question is at the peak of your historic routing cycles, I've seen over the last 15, 20 years, the revenue growth can exceed 20% and then at the bottom of the cycle, the revenues can actually be flat to slightly negative. So with that in mind, as we go forward, I mean there are things Juniper can do differently to kind of manage the business because it is cyclical, manage just kind of the mix and the cash flow of the margins. And then a question for you Bob. In security, I've been you've been -- Juniper has been trying to stabilize the business for some time. You have $2.8 billion in cash, which incidentally is the check that Cisco wrote this morning to acquire Sourcefire. So with that in mind, how do you accelerate change for the security division with the limited resources? And is the stability in the business worth all the investments that you're making or is it better to kind of reallocate those resources and refocus on wireless, core routers or PTX, which is your traditional areas of strength. Robyn M. Denholm: Okay, so I think Kevin will start on the third question, and then I can add on that, and then we'll hand it over to Bob. Kevin R. Johnson: Yes, in terms of -- Mark, in terms of your question of managing the mix, one of the strategies that we think helps us do that is customer diversification. So even within service providers, the fact that we focus on not only just carriers, which is our traditional customer base of wireline and wireless, but also cable and content providers, that helps us manage the mix. And then the fact that we've got strong enterprise business as well that has a wider base of customers and industry verticals. So if one industry vertical is maybe in a down cycle, there's another one that's in an up cycle. So that diversification of customers across service provider and enterprise is really a core part of our strategy to diversify that mix. Likewise then, focusing on 3 products or 3 business areas, routing, switching and security, is also a key part of what we do to try and manage that mix. And certainly, when you see cycles where we have good performance in 2 of those business areas and maybe weaker performance in a third, that helps balance things out. Certainly the challenge is when you run R&D and the type of projects that Bob and Rami oversee, many of those are certainly multi-quarter, but in many cases, multi-year projects. When we take decisions in terms of silicon roadmaps and major systems roadmaps and things we're doing, those become a longer cycle. And so the investment that you have to make has to endure both times when revenue is up and revenue is down. But at the core of it, we try and diversify the customer base, both in service provider and enterprise, to ensure that we don't have as much, I guess, variability in the overall op margins or profitability of the company. Robyn M. Denholm: Yes, and I would also add, obviously, over this last cycle, we have not only learned a lot, but also ensured that we've taken the right actions from a cost perspective and made sure that we're continuing to focus on that side through even a spending cycle that we say with the service providers and also improved demand on the enterprise side, as well. So on the security side, would you want to comment, Rob? Robert L. Muglia: Sure, I'll make a couple of comments, Mark. Thanks for your question. On the security side, we talked about accelerating change and when I think about that, first, we've been very open for some number of quarters that would have some challenges in our security business, in particular, the challenges as we executed the transition from the NetScreen product line to screen -- over to the SRX and some of the issues we've had with enterprise. When I think about accelerating changes, I sort of divide that into 3 sections. There's the side of organic things we can do inside our R&D organization, there's some inorganic things that one might do and then there's the go-to-market side. And we're really focused, at some level, on all 3. Most of our change over the last 18 months has been on our organic R&D where we've made major changes in our management team, in our engineering team and have really, really reinvigorated the work we're doing there. We are attracting talent and are continuing to attract talent in the security business, including in the last quarter where we've added some great new talent that we think will pay dividend in the long run. On the inorganic side, we haven't done anything at the $2.8 billion level and we don't really anticipate anything like that. But we have done some very strategic tuck-in acquisitions that have helped us to move into some new business areas and you saw that with Mykonos, you saw that with the work that we've done on DDoS Secure. We think those will pay dividends in the long run. And as I mentioned in my prepared scripts, we're doing some things in terms of our go-to-market efforts on the enterprise and in security to really help accelerate our ability to move forward in that business, so there's changes we're making there. In terms of your comments about is it worth it, we pretty strongly believe the answer to that is, yes, and I'll give you a couple of reasons for that. One is that when we look at the growth of sorts of businesses that are growing security, overall, we believe will continue to be a growing business in the years to come. When we see the challenges that our customers are facing with hackers attacking their data center environment and stealing their intellectual property, those threats are going to continue to increase, so the sophistication of the ability to defend against that is also going to increase, and we believe we're very well positioned to actually capitalize on that and actually have a leadership position there within WebApp Secure and Spotlight. The other thing I'll point out is just that we see great complements with our security business and our networking business. That's true in the systems side. It is also true as we move in the software side with SDN. And as I mentioned, quite a few of our trials of the SDN work we're doing brings together both our networking side as well as our security side, and we think that better positions us competitively in the market.

Thank you. That is all the time we have today. We'd like to thank everyone for your great questions and for your participation on today's call. Thank you. Bye, bye.

Ladies and gentlemen, this does conclude today's teleconference. You may disconnect your lines at this time. Thank you for your participation.