Greetings. Welcome to the Juniper Networks Q2 2023 Financial Results Conference Call. At this time, all participants are in a listen-only mode. A question-and-answer session will follow the formal presentation. [Operator Instructions] Please note this conference is being recorded. I will now turn the conference over to your host, Jess Lubert. You may begin.

Thank you, operator. Good afternoon, and welcome to our second quarter 2023 conference call. Joining me today are Rami Rahim, Chief Executive Officer, and Ken Miller, Chief Financial Officer. Today's call contains certain forward-looking statements based on our current expectations. These statements are subject to risks and uncertainties and actual results might differ materially. These risks are discussed in our most recent 10-Q, the press release furnished with our 8-K filed today, the CFO commentary posted on the Investor Relations portion of our website today, and in our other SEC filings. Our forward-looking statements speak only as of today and Juniper undertakes no obligation to update any forward-looking statements. Our discussion today will include non-GAAP financial results. Reconciliation information can be found on the Investor Relations section of our website under Financial Reports. Commentary on why we consider non-GAAP information a useful view of the Company's financial results is included in today's press release. Following our prepared remarks, we will take questions. We ask that you please limit yourself to one question, so that as many people as possible who would like to ask a question have a chance. With that, I will now hand the call over to Rami.

Good afternoon, everyone, and thank you for joining us on today's call to discuss our Q2 2023 results. We delivered better-than-expected results during the second quarter, with total revenue of $1.430 billion, growing 13% year-over-year and exceeding the midpoint of our guidance. Product revenue grew by 15% year-over-year and we saw year-over-year growth across all geographies. Profitability was also strong in Q2, as our non-GAAP gross and operating margin, both exceeded expectations, resulting in non-GAAP earnings per share of $0.58, towards the high-end of our quarterly guidance range. Our Q2 results reflect strong execution by our team, as well as the improvements we're seeing in the availability of supply. We remain confident in our positioning from a technology perspective and our ability to capitalize, as our customers build their networks for the next decade, which we believe will leverage AI powered software automation, an area where we have invested meaningfully over the last few years. We believe these investments, along with our go-to-market focus, will enable us to not only deliver sustained top-line growth, but also improve profitability, even in a challenged end market environment. With respect to demand, total product orders grew nearly double-digits on a sequential basis. And while the year-over-year rate of decline improved relative to last quarter, it was still meaningful due to the strong sequential performance in Q2 2022. While we are continuing to see positive momentum in our Enterprise business, we experienced weaker-than-expected trends with our Cloud and Service Provider customers, which we believe is due to the timing of project and the digestion of prior purchases. Despite these trends, we continue to expect the decline in orders to moderate further over the next few quarters and return to year-over-year growth potentially as soon as Q4 of this year. From a vertical perspective, I remain extremely encouraged by the momentum we're seeing in our Enterprise business, which delivered record revenue results and accounted for more than 45% of our total revenue, representing both our largest and fastest growing vertical for a third consecutive quarter. Not only did our Enterprise revenue grow by nearly 40% year-over-year in the Q2 time frame, our Enterprise product orders also saw healthy year-over-year growth despite a 20%-plus comp in the year-ago period. Importantly, new logos grew by more than 30% year-over-year, which we view as an important forward indicator, given the opportunity to expand after landing many of these accounts. Not to be overlooked, deal registration through the channel and commercial orders both grew by more than 40% year-over-year, which we think speaks to the differentiation of our products and our ability to capture share. Within the Enterprise, our campus and branch business had another record quarter in Q2 with our AI-driven Enterprise revenue growing more than 60% year-over-year. Customers are recognizing Juniper's clear and defensible leadership when it comes to AI driven operations delivered via a modern microservices cloud. While the rest of the industry continues to talk, we have real AI solutions that deliver real results, including a 90% reduction in worldwide trouble ticket at a global software company, 85% fewer store visits by IT at a multinational retailer and the fastest branch network rollout in the history of a national mobile operator. Revenue from the Mistified segment of our business, which are products driven by Mist AI, had a record quarter, growing by nearly a 100% year-over-year in the Q2 time frame, with orders growing by nearly 40% year-over-year. Strength was broad across the portfolio with a record wireless, wired and SD-WAN revenue in the quarter, as well as record full stack wins where customers purchase several of these campus and branch products together. We view momentum with these full stack wins as a positive forward-looking indicator, given our belief that for every dollar of wireless, there is $2 to $3 of wired switching and additional SD-WAN opportunity. Marquee new AI-driven Enterprise customers this quarter include a Fortune 10 technology company, a Fortune 50 financial institution, an American supercenter chain, the UK's largest cycling retailer and a multinational manufacturing company. We introduced several new innovations to the Juniper Mist portfolio this past quarter, including the industry's first AI-driven, Cloud-based network Access Assurance solution, which we believe has the ability to revolutionize a very dated NAC industry. In addition, we expanded our AIOps leadership by integrating the Marvis Virtual Network Assistant with ChatGPT for enhanced knowledge-based queries using large language models. And we integrated with a leading Internet collaboration platform for superior video performance. Our Enterprise datacenter business also performed well in Q2, with Apstra reporting a record quarter, both from a revenue and an orders perspective. The Apstra pipeline continues to grow with new logos more than doubling year-over-year for a second consecutive quarter, and we continue to see strong hardware pull-through for every dollar of software, which we view as a positive indicator for our Enterprise data center prospects. Our automation driven data center revenue in the Enterprise posted a record quarter, primarily due to the increasing adoption of Apstra. New Cloud-ready data center wins this quarter include a Fortune 250 financial services firm, a large US restaurant chain and a large US government agency. The performance of our Enterprise business shows our diversification strategy is working and given our level of portfolio differentiation, balanced against our relatively modest share in the large markets where we compete, I expect us to grow both our Enterprise revenue and orders during the year, even in a more challenged macro-environment. As highlighted over the last few quarters, we continue to see accounts across each of our customer verticals more closely scrutinizing budgets and project deployment timelines due to the macro uncertainties that are happening around the world. This is proving to be particularly true in the Cloud where we're seeing more customers digesting prior purchases and pushing projects to future periods. While these dynamics are likely to pressure our Cloud business for the next few quarters, we remain optimistic regarding our longer-term growth prospects in the Cloud, given our strong wide area footprint, the rapid traffic growth that continues in many of these customer environment and the opportunity to capitalize on the adoption of large language model and the build-out of AI clusters. To this last point, we expect AI adoption to drive a meaningful uptick in traffic growth that is likely to benefit our Cloud wide area footprint over time. However, we also see an attractive data center opportunity emerging, where we believe the performance and power efficiency of our custom silicon, the congestion management capabilities embedded within our Junos operating system and our support for technologies such as RDMA networking will position us well to capture share, particularly with Cloud Major accounts that are likely to bet on Ethernet as the protocol of choice to support their AI cluster investment. In fact, we've already begun to see successes in non-hyperscaler accounts. Our Service Provider business performed as expected in Q2, but moderated on both a sequential and year-over-year basis, following the strong shipments we experienced in the Q1 time frame. We expect this business to remain lumpy going forward, as the continued momentum we're seeing with respect to 400 gig deployments particularly with some of our larger Tier 1 customers is being offset by incremental weakness with Tier 2 and Tier 3 carriers that are being impacted by the softer macro-environment. Despite these headwinds, we remain encouraged by the momentum we're seeing in our cloud metro portfolio, where our new ACX7K platform experienced a record quarter, both from a revenue and in orders perspective, and the pipeline of opportunities remains strong. We expect this business to build through the remainder of the year and become more material to revenue in 2024 and beyond. In summary, I remain confident in our strategy and optimistic regarding our long-term growth prospects. My enthusiasm is fueled by our continued Enterprise momentum and the attractive longer-term opportunities we continue to see in the Cloud, as well as the SP metro opportunity. However, given the digestion of prior purchases and the uncertain timing of customer deployments, particularly amongst some of our larger Cloud customers, we have less visibility and our revenue results are likely to be pressured over the next few quarters. Based on these dynamics, we are reducing our full-year revenue growth forecast. We remain committed to delivering improved profitability and still expect to deliver greater than 100 basis points of operating margin improvement in 2023. I will now turn the call over to Ken, who will discuss our quarterly financial results in more detail.

Thank you, Rami, and good afternoon, everyone. I will start by discussing our second quarter results, then provide some color on our outlook. We ended the second quarter of 2023 with $1.430 billion in revenue, above the midpoint of our guidance and up 13% year-over-year. We delivered non-GAAP diluted earnings per share of $0.58, towards the high-end of our guidance range driven by higher-than-expected revenue and gross margin. From a customer solutions perspective, on a year-over-year basis, AI-driven Enterprise led the way with revenue growth of 63%. Automated WAN solutions grew 3% and cloud-ready data center revenue was flat. Looking at our revenue by vertical, on a year-over-year basis, Enterprise increased 38%; Service Provider increased 1% and Cloud decreased 6%. Total software and related services revenue was $318 million, which was an increase of 49% year-over-year. ARR was $319 million, and grew 37% year-over-year. We were pleased to see nearly 60% year-over-year growth in our SaaS and software license subscription portion of our deferred revenue. We remain confident in our outlook for total software and ARR growth. Total security revenue was $168 million, up 6% year-over-year, due to the timing of shipments related to improved supply. In reviewing our top 10 customers for the quarter, six were Cloud, three were Service Provider and one was in Enterprise. Our top 10 customers accounted for 27% of our total revenue as compared to 34% in the second quarter of 2022. Non-GAAP gross margin was 58.3%, which was above the midpoint of guidance, primarily driven by favorable software revenue mix and higher revenue volume, partially offset by higher inventory-related expenses. Non-GAAP operating expenses increased 10% year-over-year, primarily due to headcount-related costs, but were flat sequentially. Non-GAAP operating margin was 16.9% for the quarter, which was above our expectations, driven by higher revenue and better-than-expected gross margin. Cash flows from operations were $343 million, which was benefited by approximately $200 million in deferred federal tax payments that will be paid later in the year. We paid $70 million in dividends, reflecting a quarterly dividend of $0.22 per share. We also repurchased $120 million worth of shares in the quarter. We exited the second quarter of 2023 with total cash, cash equivalents and investments of $1.3 billion. Despite the challenging macro-environment, our results for Q2 were very strong, and I'm pleased with our team's dedication and commitment. Now I would like to provide some color on our guidance, which you can find detailed in the CFO commentary available on our Investor Relations website. For the third quarter, we expect to see continued weakness in orders, particularly with our Cloud and, to a lesser extent, our Service Provider customers. We believe the softness in orders is largely attributable to customer digestion of previously placed orders and certain projects being pushed to future periods. We expect the macro-environment to remain challenged which may continue to impact customer spending. These factors are negatively impacting our revenue expectations. Non-GAAP gross margin is expected to modestly increase sequentially in the third quarter 2023. This forecast assumes supply-chain related cost improve but remain elevated relative to pre-pandemic levels. We will continue to manage non-GAAP operating expenses prudently and expect a sequential decline in the third quarter as compared to the second quarter of 2023. Turning to our expectations for the rest of 2023. As mentioned previously, we experienced weaker-than-expected order activity in the second-quarter, particularly with our Cloud and to a lesser extent our Service Provider customers. We expect this weakness to continue into the second-half of the year. As such, we are amending our full-year revenue guidance to approximately 5% to 6% growth. However, we are raising our non-GAAP gross margin guidance from approximately 58% to greater than 58% on a full-year basis, primarily driven by anticipated improvements in supply chain costs. Additionally, we still expect to deliver non-GAAP operating margin expansion of greater than 100 basis points as the benefit from higher non-GAAP gross margin and prudent non-GAAP operating expense management should offset the impact of lower revenue. Non-GAAP earnings per share is expected to grow double digits in 2023. Our long-term financial objectives have not changed. We plan to deliver sustainable revenue growth, improved operating margin and earnings expansion over time. Finally. I am pleased to announce we have declared a quarterly cash dividend of $0.22 per share to be paid this quarter to stockholders of record. In closing, I would like to thank the Juniper team for their continued dedication and commitment to Juniper’s success, especially in this dynamic environment. Now I'd like to open the call for questions.

Thank you. [Operator Instructions] Our first question is from Simon Leopold with Raymond James. Please proceed.

Thanks for taking the question. I wanted to discuss the weakness you called out on the Cloud orders. I'm wondering if you see this as an issue actually stemming from AI projects pulling budget from the areas where Juniper has its use cases. And if so, how do you think about really the timing and extent of the eventual knock-on effect which you mentioned? So I understand AI could be a longer-term tailwind. I'm looking for a little bit more substance on that comment? Thank you.

Yeah. Thanks for the question, Simon. I think the biggest issue right now that we're facing in the Cloud provider segment is the period of digestion. It's not specific to any one customer. It's broad-based It's not even just hyperscalers. It includes the Cloud Major segment as well. To the question around AI and it being a priority, I definitely see it. There is no doubt that our Cloud customers are focusing pretty intensely on artificial intelligence, making sure that they're equipped to deal with the additional services they're going to be offering and the demand that's going to place on their network while in the short term that might have a bit of a negative impact, I think to your point in the long term, it's actually quite attractive for us for two reasons. To the extent that AI is a new killer app that's going to be offered by cloud providers to their customers, it's going to result in an increase of traffic across the board, in areas in the wide area where we have significant footprint and, of course, in the data center. And I do think as I mentioned in my prepared remarks that it could offer a nice new opportunity for us to participate in, in the data center with AI clusters being built increasingly over time with Ethernet as an alternative to what is today the technology of choice which is InfiniBand. That said, the timing to your question is -- it's difficult to call right now, but it's really in the several quarters.

Great. Thank you.

The next question is from Samik Chatterjee with JPMorgan. Please proceed.

Yep. Hi. Thanks for taking my question. I guess, I wanted to get a sense, you're trimming your revenue guide for the full year today by about $150 million or so. I mean, the incremental weakness that you're mentioning to the last quarter is Service Provider, but if you can break down sort of how much of that revision on the guide is really coming from Cloud versus Service Providers and maybe just in terms of thinking about sort of the reduction here, I think investors are going to question sort of the discussion or elevated backlog that most companies have discussed for a while just given that even with the elevated backlog, the visibility into revenue remains pretty limited. So how are you guys thinking about sort of backlog in terms of what that backlog implies going forward, particularly given that some of the macro trends are still influencing revenue on a more real time basis? Thank you.

Sure. Thanks for the question. Yeah, so for the guidance, the reduction in revenue expectation for the year is largely Cloud. Right? I would say to a to a lesser degree, Service Provider, but the vast majority of that reduction would be attributable directly to our Cloud business. On a on a year-to-date basis, our cloud business is down 10%. I think it's likely that it ends the year down double digits. And really due to the weakness we saw in orders late in Q2 as well as what we expect to see for the rest of the year. So the vast majority of that is Cloud. I believe our Service Provider business will stay within kind of our guidance range of plus or minus 2%. So that's going to be largely intact with our long-term model. And Enterprise clearly way ahead of our long-term model and I expect it to remain way ahead as we exit the year. So the revenue decline is really mostly a Cloud phenomenon. From a backlog perspective, clearly backlog is coming down as we expected it to this year. I'll tell you it came down a little bit more than we expected given that the order weakness that we called out in Q2. So backlog is coming down. However, it still remains about 3x, what we would normally expect in kind of a pre-pandemic historical level. And I expect to exit the year -- I still expect to remain elevated as we exit the year at approximately 2x kind of normal historical level. So it's coming down quickly, but it still remains quite elevated. It is going to put some pressure on 2024 revenue growth. The backlog is a tailwind in 2023. It would be a headwind to 2024 at the backlog, but we will not have as elevated backlog entering next year as we did this year. It's going to require us to have a recovery in bookings and orders. The good news is I expect that we will have a recovery in orders. I expect next year to be a strong bookings year for all of our verticals as we kind of go through this backlog digestion period and react to the current lead time. So this year is obviously putting a lot of pressure on bookings. It's resulting in a declining backlog. I think next year would be a little more normal, but bookings will drive revenue growth.

Thank you.

Up next, we have Tim Long with Barclays. Please proceed. Tim, your line is live.

Thank you. Yeah. I wanted to touch on the Enterprise business Obviously, good numbers there, still a lot of new customers. Could you talk a little bit about kind of where we're at with kind of win rates, deal sizes, kind of cross selling if you could just give us a little more color on what's driving that? And related, if you can just touch on -- we're seeing macro impact a lot of other places. It looks like it's starting or going to start hit that Enterprise networking area soon. How do you view the success you've had in the market share gain in a backdrop where we might be seeing more pressure overall on the industry? Thank you.

Okay. Thanks for the question, Tim. Yeah. So I'm obviously very pleased with the performance of our Enterprise business. And really, this is in the face of macro-related headwinds. Enterprise is now our largest, our fastest growing segment in Q2 with over 45% of total revenue. The level of differentiation that we have in our Enterprise portfolio has never been stronger. You see it in the numbers. Our AI driven Enterprise business grew at over 60%, Mistified revenue grew at over approximately 100% year-over-year and we also saw some really encouraging signs with Apstra in the data center, specifically again for the Enterprise. The question around -- really it's around durability. What happens if there were going to be sort of signs of weakness in the market? I'd just say that the Enterprise market if it was -- it's -- there are headwinds in the in the market today. Absent those headwinds, we'd actually be doing even better. We'd be posting even better results than we have right now. And as a -- just a point of consideration, take a look at how our Enterprise business performs even during the early days of the pandemic where peers in the industry were not seeing growth we were seeing growth. And that was through new logos, share taking, which is really happening right now. I'm very encouraged with the forward-looking metrics, things like 30% growth in new logos, deal registration up 40%. This is really an indication of how much opportunity our channel partners are bringing to Juniper, commercial business up at over 40%. This is -- these are signs I believe of success to come in the Enterprise. So for that reason, I'm encouraged by the business traction thus far, revenue growth and order growth this year, and it's I think fairly easy to call revenue growth for next year in the Enterprise.

Thank you very much.

The next question is from Alex Henderson with Needham. Alex, please proceed.

Great. Thank you very much. A couple of just mechanics issues. Can you talk a little bit about linearity in the quarter, pricing in the quarter? Whether the push outs were both in the Cloud and in the Service Provider. And when you talked about the scale of the 24 headwinds, can you give us some sense of what the scale is on that? Is it a 2% headwind, 4%, 6% more? And what's the split there between Enterprise, Service Provider and Cloud? Thanks.

Maybe I'll start with the 2024 view and then, Ken, I'll let you comment a little bit more about linearity and pricing. And Alex, I appreciate the question. So it's a little too early to provide guidance on 2024 all up, especially given some of the digestion that we're seeing in the cloud and to a lesser extent in the SP segment. However, I do think it is possible for us to grow revenue in 2024, and I'll just walk you through just some of the puts and takes to that. Backlog draws, Ken just mentioned in 2023 creates a pretty meaningful revenue headwind for us in 2024. So if you think about it, orders would need to accelerate from here to overcome the backlog related headwind that this year is presenting to next year. Now the good news is we do expect orders to grow and in fact to accelerate for the full year in ‘24. And in fact, we see the order growth to happen across all customer segments, in Cloud, in SP and in Enterprise. And the Enterprise, as I just mentioned, which is the bulk of our business continues to do well today both from an order and the revenue standpoint, it's very easy to call revenue growth for Enterprise next year. Cloud and SP, I’d just say it's a little bit too early for us to call right now. Orders will accelerate. We don't know if they're going to accelerate fast enough to overcome the backlog related headwinds for next year. So the summary, what I would say is, so while 2024 full year revenue growth is possible, it really is going to depend on the timing of order recovery for Cloud and SP. And I would say, I'm sure Ken would say this if I don't, we're committed to both margin and EPS growth next year in 2024 irrespective of what Cloud and SP does.

Yeah. From a linearity perspective, I'm actually going to answer a couple of things. First from our Q2 orders, the weakness in orders really showed up late in the quarter. So that was something that happened late in the quarter. And our current visibility is that the second half of the year are also going to be down as compared to our previous expectations in the Cloud and to a lesser degree, Service Providers. So it really did kind of show up late in the quarter. I do want to touch on linearity for next year as well. I do believe it's going to be more of a back-end loaded year, a little bit different than, say, this year. So the first half of next year is going to be, I would say, particularly challenged from a growth opportunity perspective. And full year growth is too early to call. It is possible, but there's a lot of moving parts there. But I would set up a back-end loaded year, given the current situation we see and the weakness we see now is likely to bleed into the first half of next year. The other question you asked a little bit on, Alex, was pricing. We really didn't notice a material change in pricing activity. It's always been a very competitive pricing environment. It's kind of deal to deal, hand-to-hand combat in many ways. So it's really nothing to call out from a more holistic point of view on the pricing side.

Great. Thanks.

The next question is from David Vogt with UBS. Please proceed.

Great. Thanks guys for taking my question. Maybe, Rami, can I just follow up on the Cloud digestion sort of train of thought here. And how are you thinking about where Juniper kind of fits in or slots in as either the hyperscalers or the major cloud providers really start to accelerate maybe CapEx next year in terms of where you slot in maybe from a timing perspective? And along those lines, are you seeing some of the order digestion or maybe pushouts being driven by maybe a reallocation of CapEx from Cloud customers, whether it's the hyperscalers or main cloud providers? Are they spending more on GPU-enabled servers and compute and maybe that's putting pressure on other parts of network spend? And would you just kind of give some thoughts and color on what you're hearing from your customers at this point? Thanks.

Yeah, I appreciate the question, David. I think the biggest factor right now that's impacting our Cloud business is digestion. It's just unprecedented order growth over the last couple of years now being followed by a period where they need to work through the products that they've ordered, they need to receive them, they need to deploy them, and then revenue will catch up. Orders -- or recovery in revenue will catch up afterwards. As I mentioned earlier, there definitely is a lot of attention and maybe even some diversion of capital investments to AI clusters, GPUs, that sort of thing. And yes, I think that can have a short-term negative impact, not the biggest impact. Again, it's not as big, I would say, nor as near as big as that -- just of digestion. But the long-term impact of that AI cluster investment, in my view, is a net positive. And this is true for areas of the Cloud provider network where we have significant market share in hyperscale WAN, for example, but it also presents opportunities for Juniper to participate in a AI cluster Ethernet networking, which I believe becomes more dominant in time.

Great. Thanks, Rami.

My pleasure.

Up next, we have Karl Ackerman with BNP Paribas. Please proceed.

Yes. Good afternoon, gentlemen. I was hoping to tie in or have a follow-up question with regard to some of the comments you asked -- or discussed earlier. One is, I understand that pricing has not really changed yet. At the same time, you indicated that orders were toward the end of the quarter. So I guess are Service Provider and Cloud customers pushing back on price towards the end of the quarter? Is that some of the reason for your cautiousness? And then that's a clarification. I have a short question as well.

Yeah. So just to clarify, I was mentioning that at the end of the quarter orders, the weakness showed up late in the quarter. So actually, our orders were weaker at the end of the quarter than we would have expected as we started the quarter. So linearity was not back-end loaded. In fact, it didn't come in as strong as we thought it would over the last few weeks of the quarter from a bookings perspective.

I see. And I was hoping you can just maybe address why hardware maintenance and professional services have moderated the last two quarters despite the improving strength in Enterprise? I suppose, should we expect that to flip going forward? Thank you.

Yeah. So our hardware and maintenance business really is an installed base business. So although current period in revenue and recent period revenue obviously matters, it really is about the installed base. And we have seen significant strength in that business over the last several years, where it's been growing pretty routinely, maybe not as much as product when product is growing well, but it grows -- it still grows even when products has a down period. So it's a pretty stable business that's growing in line with our expectations, to be honest with you. There is a shift happening as well to more software. So when you isolate just hardware maintenance and support, you are kind of underestimating that the true power of our services business, which is on the software side, which shows up in a different line item.

Thank you.

Up next is George Notter with Jefferies. Your line is live.

Hi, guys. Thanks very much. If I go back to last quarter, if I remember correctly, you guys were also pointing to Cloud customers going through a period of inventory digestion. And I guess it seems like it really intensified over the last three months. I guess my question for you is that, is that the right observation? And do you get any commentary from them about why that's intensified? And then also, I just want to make sure this isn't a market share issue. Any sense for market share or your participation share-wise in some of these accounts? Thanks.

Yeah. So to answer the second part of the question first, no, we do not believe this is a market share issue. Just keep in mind that in hyperscale, our dominant footprint is in the wide area. So it is very much around sort of capital intensity, specifically for wide area investment and so on. The first part of your question, the answer is yes. I mean, it did intensify. We obviously had some expectation of how long the period of digestion would take, how long it would take to deploy new projects and it has elongated. Having said that, it's important to say that projects that we have been working on and competing for remain intact. We have not seen Cloud providers cancel projects. 400-gig deployments in the wide area, for example, remain incredibly important, especially with cloud providers having to invest in new AI capabilities. Ultimately, that's going to generate more demand, traffic growth on their network across every use case. And the projects that were in place might have moved out in time, but they have not been eliminated. And it continues to be an intense area of focus for us at Juniper.

Got it. And then just a quick follow-on. I know you guys are working on new silicon for the PTX. I'm just curious about where that is. That's obviously, I think, an important new product delivery for Cloud customers. Just wondering where that is.

Yes. So I mean, it's looking great, honestly. I mean the level of innovation and just the sophistication of the silicon technology that we've developed, and the latest express silicon looks just absolutely remarkable. Customers that have had an early view of what we've developed, are incredibly excited about it. I believe from a market fit standpoint, it absolutely is addressing some of these big cloud provider use cases that I just discussed, and we're getting close. So stay tuned.

Thank you.

The next question is from Mike Ng with Goldman Sachs. Please proceed, Mike.

Hey, good afternoon. Thank you for the question. I just had some follow-ups to your comments on Samik's question on backlog. So you started the year with over $2 billion in the backlog. Are you still expecting to exit the year with backlog levels around $500 million to $1 billion? Any comments on where those net backlog reduction sat in 2023 by customer vertical? And then I just have a quick follow-up. Thanks.

Yeah. So we still expect to exit the year on an elevated level, about 2x normal. And we define normal historically as -- we were $400 million plus or minus a bit for quite a period of time. So I'm talking about maybe the middle to the high end of that $500 million to $1 billion range that you talked about, $800 million, plus or minus $100 million or $200 million would be where I'd expect our backlog to exit for the end of the year. So still elevated. To date, it's been coming down fairly proportionately across all verticals. We entered the year with a pretty balanced backlog as compared to what our revenue is. That said, I think we'll exit the year with a little bit more in the Enterprise side, and we'll probably burn through the Cloud backlog at a little faster rate than the other verticals, given some of the weakness that we've mentioned on the new orders.

Great. Thank you. That's really helpful. And just a housekeeping question. Orders being up nearly double digits sequentially, my model kind of suggests that implies down mid-20s on a year-over-year basis. Is that kind of in the ballpark?

Yeah, we did see a modest improvement from last quarter on a year-over-year basis, and we expect modest -- we expect improvement going forward. And we could return to growth as soon as Q4. So yes, we are -- we did see some modest improvement. So that's in line, I think, with kind of where you are.

All right, great. Thank you.

The next question is from Meta Marshall with Morgan Stanley. Meta, your line is live.

Great. Thanks. I noted that you guys were saying that there was some elongation with maybe Tier 2 and Tier 3 Service Providers and Cloud. Those have been areas where you had a lot of new customer wins in the last year. And so I guess I'm just trying to get a sense of -- is this some of the new wins that you had over the last year just taking longer to implement or is it kind of broader across all of the customers within that area? And then the second question is just if you could give a sense of kind of what the overhang on gross margins is still from kind of expedite fees or kind of supply chain costs? Thanks.

Okay. Thanks, Meta. So the period of elongation has been mostly at Cloud, to a lesser extent in SP. You are right, we have won, we've secured new projects over the last year or two that we have mentioned on these earnings calls. And even there, where we have started deployments, some of the deployments have just moved out in time. So it applies to areas where we have existing footprint and net new wins where we've started to see some benefit, but maybe not as much benefit as we would have otherwise enjoyed had these -- had this elongation not taken place. And again, it's mostly a Cloud phenomenon to a lesser extent, in the SP. And then the second part of the question, I'll leave it to Ken on the gross margin.

Yeah. On the gross margin side, we are starting to see some recovery in those costs. I would point to logistics, in particular, that we're seeing some recovery there as well as some of the expedite fees. I still think there's 70 basis points, maybe a little more, maybe a little less, impact to supply constraint costs. So there's still opportunity for it to get better. The other thing I would add is we did call out -- I called out in my prepared remarks, we are seeing some higher inventory-related expenses that were in Q2, and I expect some of those to persist into the second half, but that's been factored into the guidance that we provided on gross margin.

Great. Thanks.

The next question is from Atif Malik with Citi. Atif, your line is live.

Hi, thank you for taking my question. And good to see orders growing double digits sequentially. I have a question for Ken, and I apologize, I'm a bit new to covering Juniper. Ken, why provide a full year outlook given Cloud and Service weakness? Is this because of supply chain tightness that you guys saw during COVID, which stretch your lead times? I'm just trying to understand the rationale behind providing the outlook for this year and even commentary into next year. Thank you.

Yes. So with the commentary next year, just to be clear, we are not providing an outlook for 2024. We're trying to provide you with the puts and takes, right? And we do think Enterprise -- that is the one outlook we did provide. We do expect to grow Enterprise next year given the strength of our products and the differentiation we enjoy there. But Cloud and SP, way too early to call, therefore, too early to call the aggregate number. For this year, the rest of it really is just about Q3, what we're providing guidance for and really Q4. And we did provide a range, 5% to 6% for the full year. I do think that's -- based on current visibility, we feel comfortable with that guidance. We still have the elevated backlog that I mentioned earlier. That really does drive a good chunk of the revenue that we're going to recognize for the rest of the year. So it's just really the visibility we have now, we felt comfortable with that type of a range given where we are in the year.

Okay. The next question is from James Fish with Piper Sandler. James, please proceed.

Hey guys. Thanks for the question. Look, I know you guys are highlighting digestion and also highlighted Ethernet as the kind of the main protocol over InfiniBand. But Rami, for you, just any updated view why Ethernet will be best for AI workflows? And secondly, given the increased networking demand with AI, is this part of the reason for the optimism for orders to reaccelerate in 2024? And if so, why do you think Juniper's portfolio is best positioned to handle those AI workloads?

Okay. So the optimism, I would say, is not just rooted in Ethernet being sort of the technology of choice for AI clusters, that's a component of it. But the optimism really is around the fact that traffic growth in Cloud providers continues to grow. Digestion can't happen forever. Eventually, they're going to deplete through their inventory and they're going to -- and Cloud providers are going to need to invest in their networks. To the extent that they invest in the wide area, we will benefit from that just based on our footprint. And then in the data center, as you know, we've got a very good footprint in the top 10 and on the broader cloud majors. And I would add to that, and it's in addition that AI clusters moving to Ethernet would present a net new opportunity for us. And to your question about why Ethernet and not InfiniBand, I mean, obviously, today, most AI clusters are built with InfiniBand. In the future, however, Ethernet present has some distinct advantages. From an economic standpoint, it's got a broader ecosystem. So it's not just Juniper. There will be a number of players. They're going to be innovating in the Ethernet space to capture the requirements of AI clusters. That means from an economic standpoint, customers that want to deploy large language model type clusters are going to benefit economically from moving to Ethernet. But I want to just highlight that this does not happen overnight. This is going to be something that will take several quarters, likely over a year or so, in order for Ethernet to start having a meaningful volume of business tied to AI cluster networking.

Helpful. And if I could just follow up there on your comments. Any sense as to how much excess inventory the Cloud and Service Providers are holding versus normal? Is it more the inventory issue is on the Cloud side with kind of the deployment issue on the Service Provider side? Or is it kind of the same theme across both? Thanks guys.

Well, it's mostly a Cloud provider thing. It's to a lesser extent in the Service Provider space Obviously, they're sitting on elevated inventory. When we talk about digestion, really, it means that they are drawing down that inventory and deploying it into their networks. I think your question is really getting to the timing. And we don't know with certainty at this point in time. I can tell you there's going to be several quarters before, as they draw down inventory, do that deployment and they start to see the acceleration -- and we start to see the acceleration of orders again.

Okay. The next question is from Tal Liani with Bank of America. Please proceed.

Yes. Hi, guys. Sorry to ask third question on the same topic. I just wanted to get the numbers right. Last quarter, you said that backlog was roughly $1.7 billion and the drawdown was about $350 million. This time -- and in the past, you said that the normal backlog is $300 million to $400 million. So this time, you're saying 3 times the backlog. So if I take the $400 million, at high end, it's $1.2 billion, and it means that the backlog drawdown went up from $350 million to $500 million sequentially. But then you're saying that by the end of the year, it will be $800 million to $1 billion. So that means that the -- you expect the backlog drawdown to slow down materially in the next two quarters. Just want to make sure that the numbers that we have are correct? Or if you can give us -- be a little bit more specific with that because I'm sure everyone is doing the same math to understand the backlog drawdown. Thanks.

Yeah, let me add some specificity to the backlog. So we have talked about we ended the last quarter Q2 at approximately 3 times number of levels. We're between $1.2 billion and $1.3 billion, right? That would be our current backlog position between $1.2 billion and $1.3 billion. And I do still expect to exit the year at approximately 2 times. I'm using $400 million as my average. So that means approximately $800 million. So currently, $1.2 billion to $1.3 billion, expect to exit the year at approximately $800 million. That does presume a bit of a slowdown in the second half as compared to the first half. But that's at -- that's obviously based on our expectations of shipments and bookings for the rest of the year.

Got it. on the same topic, but differently, if the weakness is coming -- the order weakness is coming mainly from Cloud and a little bit less than in Service Providers. And on the other hand your Enterprise is very strong. So does it mean that the backlog drawdown is much more meaningful as a percentage of your revenues in these verticals? So I'm just trying to understand if we should assume that the backlog drawdown is mostly tilted towards the Service Providers and Cloud and less to the Enterprise, given the difference in the revenue performance?

Yeah. So I do expect this year, we'll exit with elevated. I think the proportion will be more Enterprise versus how we entered the year, which was more proportional to our revenue. That said, our revenue is also starting to slag much more towards Enterprise as well. So that backlog is getting the bookings growth. So the decline in that backlog is less than the other verticals, which are actually declining on a year-over-year basis from a bookings perspective, revenue is holding up a little bit better, particularly Service Provider revenue, which is up on a first half basis. So yes, we are burning a little bit more in those two verticals as compared to our Enterprise vertical.

Got it. Great. Thank you.

We have reached the end of the question-and-answer session. This concludes today's conference, and you may disconnect your lines at this time. Thank you for your participation.