Cisco Systems, Inc. (CSCO) Earnings Call Transcript & Summary

Hello, everyone, and thanks for joining us this afternoon. I'm Sami Badri, and I lead Investor Relations here at Cisco. With me today are our Chair and CEO, Chuck Robbins, and our CFO, Mark Patterson. You can find our earnings press release and supplemental materials, which include the GAAP to non-GAAP reconciliations, posted on the Investor Relations section of our website. After we finish today, a recorded webcast and the slides will be posted there as well. During this call, we'll refer to financial results on both a GAAP and a non-GAAP basis. When we talk about product results, we'll do so in terms of revenue, and for geographic and customer results, we'll use product orders, unless we say otherwise. Every comparison we make will be year over year. I want to note that today's remarks will contain forward-looking statements, among them our guidance for the second quarter and for fiscal year 2026. Those statements carry risks and uncertainties that are described in our SEC filings, in particular our most recent 10-K report, which lays out the key risk factors that could lead actual results to differ materially from what our forward-looking statements contain. Regarding guidance, please also refer to the slides and the press release that go along with this call for more information. Cisco won't be commenting on its financial guidance during the quarter unless that happens through an explicit public disclosure. With that, I'll hand things over to Chuck.

Thank you, Sami. I appreciate everyone being here today. Fiscal 2026 got off to a great start for us, as both Q1 revenue and earnings per share exceeded the upper bound of the guidance ranges we had provided; we posted a record for Q1 revenue, which keeps Cisco on pace for our best year ever, consistent with what our full-year guidance indicates. During Q1, total revenue rose 8% compared with a year earlier, while product revenue climbed 10%, fueled by powerful demand for our AI infrastructure offerings and campus networking solutions. The combination of our robust top-line results, operating efficiencies, and strong execution from our teams drove non-GAAP EPS up 10%, as we keep expanding earnings at a quicker pace than revenue; we produced healthy margins and cash flows, which let us give back $3.6 billion of capital to shareholders via dividends and share repurchases, equal to 125% of free cash flow in Q1. On top of that, we achieved healthy growth in annualized recurring revenue and remaining performance obligations, and both of those keep offering a solid base for our performance in FY 2026 and afterward. Cisco's impressive beginning to fiscal 2026 demonstrates how essential secure networking is and how powerful our portfolio is as organizations seek to roll out AI throughout their operations. That being said, we recognize plenty of customers still have considerable work ahead of them to make sure their networking infrastructure is modern, scalable, and secure enough to serve their AI objectives. Based on our 2025 Global AI Readiness Index, merely one-third of organizations believe their IT infrastructure is capable of handling what their planned AI projects will require, and that represents an enormous opportunity for Cisco. Thanks to our networking portfolio that leads the industry, enabled by Silicon One AI-native security solutions and operating systems, we are positioned well right now to supply the essential infrastructure for the AI era. Let me now offer my thoughts on the robust demand we observed in Q1. On the whole, total product orders rose 13% compared with a year earlier, with increases across every geography and customer market. Enterprise product orders climbed 4% year-over-year in Q1, building on mid-teens growth excluding Splunk from a year ago, with particular strength in campus switching and wireless solutions. Public sector orders rose 12% year-over-year, with gains across every geography and cohort, U.S. Federal included. Product orders from service provider and cloud customers remained extremely robust, rising 45% year-over-year, powered by high double-digit order growth and hyperscalers, even against a difficult triple-digit growth comparison from Q1 of FY 2025. Demand from Telco customers was robust as well in Q1, with orders increasing over 25% year-over-year. Now, a bit of detail on demand from the product angle. Networking product orders sped up to high teens growth in Q1, which was the fifth quarter in a row of double-digit growth, powered by hyperscale infrastructure, enterprise routing, campus switching, wireless, industrial IoT, and servers. Across our campus networking portfolio, demand is extremely strong for switching, routing, and wireless products, which shows that enterprise customers are putting money into the connectivity required for AI deployments. With older Catalyst switch generations such as the 4K and 6K approaching the end of support, demand keeps building for our Cat 9K lineup. On top of that, every one of our next-gen offerings — smart switches, secure routers, and Wi-Fi 7 wireless products — is ramping more quickly than we saw in earlier launches. This opens the door to a refresh opportunity spanning multiple years and multiple billions of dollars. Our industrial IoT portfolio is likewise showing steady gains, including new ruggedized equipment, with orders up over 25% year-over-year in Q1; we look for that demand to climb further, fueled by manufacturing onshoring into the United States, more AI workloads moving to the network edge, and physical AI beginning to emerge. AI infrastructure orders booked from hyperscalers during Q1 came to $1.3 billion, split evenly between Silicon One systems and optics, which signals a meaningful step-up in growth and shows the strength we bring to sophisticated AI use cases. From hyperscalers, we anticipate recognizing about $3 billion of AI infrastructure revenue in fiscal year 2026. Since these hyperscale customers want to stretch AI clusters throughout their infrastructure, we're seeing strong appetite for Acacia's coherent pluggable optics, which lead the market and deliver major cost and power savings. Every hyperscaler now buys these products. During Q1, we also introduced our newest Cisco 8223 router, built on our Silicon One P200 chip. This 51.2 Tb per second fixed Ethernet routing system, first to market, is engineered for the heavy AI workload traffic flowing between data centers. Thanks to Silicon One's scalability, power efficiency, and programmability — all unmatched — we can deliver performance and speed across data centers that previously would only have been achievable inside a single data center using a switching infrastructure. Silicon One demand keeps rising, and we expect our one millionth chip to ship in Q2 of fiscal year 2026. Orders for AI use cases outside hyperscaler training are gaining momentum too, with data center system orders, including switching and compute, up double digits in Q1 as customers get their networks ready for inferencing and agentic workflows. Across Sovereign, Neocloud, and Enterprise customers, we're seeing a pipeline that keeps growing beyond $2 billion for our high-performance networking products. To seize this opportunity, we keep advancing both our own portfolio and our strategic partnerships. We recently shared news of an expanded partnership with G42 in the UAE to power, connect, and secure G42's large-scale AI clusters featuring AMD GPUs. Other strategic partnerships in the region, among them Humane and Stargate UAE, are moving forward as planned. We also introduced our Sovereign critical infrastructure portfolio for European customers, letting them operate within their own air-gapped on-prem physical environments. That portfolio covers our networking and collaboration products, bolstered by security and observability. Beyond that, Cisco announced an expanded NVIDIA partnership along with our new N9100 switch based on Spectrum X Silicon. We're now the first NVIDIA partner offering networking that complies with their cloud reference architecture. The N9100, which comes out during the back half of fiscal 2026, is going to give Sovereign and Neocloud providers the operational consistency plus flexibility they need to build and manage AI at scale. In addition, we're rolling out fresh capabilities and features for Cisco Secure AI Factory with NVIDIA, which we announced in Q3 of fiscal 2025. Through these improvements, we reinforce our dedication to high-performance, secure, and trusted AI infrastructure on a global basis. We anticipate that Cisco's AI opportunity among Sovereign, Neocloud, and enterprise customers will ramp during the latter half of fiscal 2026; turning now to security, we keep seeing order growth for the new and refreshed products we've introduced, which make up roughly one-third of our security portfolio and include Secure Access, XDR, Hypershield, AI Defense, and our refreshed firewalls. Since launch, almost 3,000 customers have bought a new product, and demand for our next-generation firewalls rose in the mid-teens during Q1; that increase was partly offset by a decline in our prior-generation platforms; we keep witnessing strong performance from Splunk, having closed one of our biggest Splunk deals ever in Q1, made possible by joint Cisco and Splunk sales engagement. Splunk's ARR and product RPO both grew at a double-digit rate, as we observed a significant shift in the way customers consumed Splunk offerings in Q1, moving toward more cloud subscriptions and fewer on-premise deals.

Cloud subscription revenue gets recognized ratably, while product revenue from on-prem deals hits the books upon delivery. That shift did weigh on security revenue growth in Q1, but it's strictly a matter of timing. In fact, we're happy to see more Splunk cloud subscriptions, since they drive greater adoption and expansion and let us bring innovation to customers faster so they can unlock value from AI. Now let me touch on a few of our recent innovations. Looking at the AI opportunity, we see customer use cases expanding across training, inferencing, and connectivity, and secure networking matters more and more as workloads shift from the data center out to end users, devices, and agents at the edge. As we said last quarter, agents are changing network traffic from predictable bursts into persistent high-intensity loads, and agentic AI queries can generate as much as 25 times the network traffic of chatbots. Rather than pulling data to and from the data center, AI workloads need models and infrastructure positioned closer to where data gets created and decisions get made, especially in industries like retail, healthcare, and manufacturing. That's why last week we introduced Cisco Unified Edge, an industry-first converged platform for the network edge that brings compute, networking, and storage together into one system. Unified Edge delivers real-time inferencing for agentic and physical AI workloads, so enterprises can deploy and manage AI at scale with confidence. We also announced Cisco Data Fabric in September, a Splunk-powered architecture that unifies and manages machine data from various sources, letting enterprises build AI models using proprietary data that previously went unused. As always, these innovations are built to advance Cisco's platform advantage, where each new technology investment compounds the value of what a customer already has. In summary, demand is strong across every customer market and geography, and opportunities are expanding as our customers power their AI use cases from the data center to the edge. We keep innovating at unprecedented scale to build AI-ready data centers, power future-proof workplaces, and create a foundation of digital resilience. Our strong performance fuels our capital allocation model, returning significant value to shareholders while positioning our business for Cisco's strongest year yet in fiscal 2026, as reflected in our guidance. Now I'll hand it over to Mark for more detail on the quarter and our outlook.

Thanks, Chuck. We kicked off the new fiscal year with a solid quarter, as revenue, operating margin, and EPS all came in above the top end of our guidance, along with healthy gross margin and operating cash flow. Total revenue for the quarter reached $14.9 billion, an increase of 8% from a year ago. Non-GAAP net income came to $4 billion, rising 9%, while non-GAAP earnings per share were $1, up 10%, which shows ongoing operating leverage as non-GAAP earnings rose more quickly than revenue. Turning to Q1 revenue in more depth, product revenue totaled $11.1 billion, up 10%, while service revenue came in at $3.8 billion, rising 2% versus the prior year. Networking stood out with 15% growth, showing strength throughout the portfolio, led by high double-digit gains in service provider routing, mostly powered by revenue tied to AI infrastructure. Data center switching and enterprise routing each added double-digit growth as well, and campus switching grew in the high single digits. Security declined 2%, driven by lower prior-generation products and a move toward cloud subscriptions in our Splunk business that Chuck mentioned, partly offset by gains in Secure Firewall, Duo, and SASE. Collaboration fell 3% on weaker devices and Webex. Observability rose 6%, mainly on ThousandEyes growth. On recurring metrics, total RPO was $42.9 billion, up 7%. Product RPO increased 10%, with the long-term piece at $11.8 billion, up 13%. Total ARR finished the quarter at $31.4 billion, a 5% increase, as product ARR grew 7%. Subscription revenue totaled $8 billion, which was 54% of Cisco's overall revenue. Software revenue overall rose 3% to $5.7 billion. Q1 product orders climbed 13% from a year earlier. Orders rose in every geographic segment: the Americas gained 16%, EMEA gained 8%, and APJC gained 13%. Orders also increased across all customer markets, with service provider and cloud up 45%, public sector up 12%, and enterprise up 4%. Total non-GAAP gross margin was 68.1%, a drop of 120 basis points year-over-year, though it landed a bit above the midpoint of our guidance range. Non-GAAP product gross margin reached 67.2%, down 170 basis points, hurt by unfavorable mix and pricing, partly offset by productivity gains. Non-GAAP services gross margin was 70.7%, an increase of 40 basis points. We remain focused on boosting profitability and maintaining financial discipline, with non-GAAP operating margin of 34.4%, exceeding the top end of our guidance range. Our non-GAAP tax rate for the quarter was 19%. On the balance sheet, we closed Q1 with $15.7 billion in total cash, cash equivalents, and investments. Operating cash flow was $3.2 billion, a decline of 12% stemming from investments aimed at meeting rising customer demand for AI infrastructure. On capital allocation, we gave back $3.6 billion to shareholders in the quarter, made up of $1.6 billion in quarterly cash dividends and $2 billion in buybacks, leaving $12.2 billion available under our repurchase program. In closing, fiscal 2026 got off to a strong beginning for us, as both revenue and profit came in above what we had anticipated, powered by robust order expansion and margins, which together show how our innovation engine can generate considerable sales growth while operating leverage boosts earnings. We stay committed to strategic investments and innovation so we can seize the meaningful growth opportunities that lie ahead. Disciplined management of spending will keep supporting this, and that potent mix keeps driving robust cash generation along with our capacity to deliver substantial value back to shareholders. Moving to guidance, keep in mind that our outlook for Q2 and for fiscal year 2026 presumes that tariffs and exemptions as they currently stand stay in effect until fiscal 2026 ends. Those assumptions are the same as in our earlier guidance, except that the China fentanyl tariff has been cut from 20% down to 10%. Going forward, you should expect that we will keep concentrating on durable growth, with financial discipline producing operating leverage and ongoing capital returns. For fiscal Q2, here is our guidance. Revenue should land between $15 billion and $15.2 billion; non-GAAP gross margin should come in between 67.5% and 68.5%. Non-GAAP operating margin should fall between 33.5% and 34.5%. Non-GAAP earnings per share should come in between $1.01 and $1.03. We are assuming a non-GAAP effective tax rate of roughly 19%. For fiscal year 2026, here is our guidance; revenue should land between $60.2 billion and $61 billion. Non-GAAP earnings per share should come in between $4.08 and $4.14. Sami, let's now move into the Q&A.

Thanks, Mark. Prior to opening things up for questions from the analysts, please remember that each of you should pose a single question along with just one follow-up when it's your turn. Operator, would you bring in the first analyst waiting?

Aaron, thanks for the question and the kind words about the quarter. I'd like to dig a bit further into those AI orders. It seems we're on track toward, if I recall correctly, $3 billion in fiscal 2026 out of the WebScale vertical. Let's perhaps begin there. When we consider the variety Cisco is encountering within the WebScale opportunity, how has that changed over time? Have you all been involved in going deeper on super spine, or perhaps scale across opportunities, within the WebScale vertical? Then, turning to the enterprise side as a follow-up, I believe the figure was a $2 billion pipeline. Per the slide deck, I think it mentions $200 million in orders during the last quarter. What's our expectation for how that develops across this fiscal year? Thank you.

Aaron, thanks. Let me clarify that the $3 billion figure we provided was a revenue figure, coming from Hyperscale AI infrastructure, during the fiscal year. I'd note that the $1.3 billion represented new orders booked during the quarter from those same customers we measured a year ago, and it's the same products we measured a year ago; it's absolutely apples to apples, above all. What we anticipate on the orders side this year is at least double the orders we booked in fiscal year 2025 from that identical customer set; we're seeing plenty of strong pipeline across the remainder of the year. On use cases, we see expansion. Naturally, we've been selling networking infrastructure beneath the training models; we've been selling scale-out. We introduced the P200-based router, which will start to tackle some scale-across opportunities. We've clearly enjoyed terrific success with our pluggable optics. Every one of the Hyperscalers is now officially a customer for our pluggable optics, so we view that as a fantastic opportunity. Not only do they plug into our products, they can also be deployed with other companies' products, whether our competitors' or white boxes. That's been positive. We've also started seeing inferencing use cases, where we're winning as well. Another thing we observed in Q1: four of the major Hyperscalers grew triple digits during the quarter, and we had four meaningful use case wins during the quarter, one from each of those four, so spanning four different Hyperscalers. The momentum with the silicon carries on. On your second question about the enterprise, what we stated was that the Neocloud, Sovereign Cloud, Enterprise pipeline, essentially for the remainder of our fiscal year, meaning the next three quarters, is above $2 billion. We booked $200 million in Q1, and that gives us incremental opportunity looking ahead, and that runs through the close of our fiscal year. Thank you.

Thank you, Aaron. Michelle, can we move to the next analyst, please?

I'd like to add on to what Aaron asked. Regarding the momentum you're observing, does the potential you see in those AI orders stem from scale-across getting stronger, or does it come purely from deeper engagement? Our take is that scale-across did get stronger over the course of the quarter. I'd just like to know whether you've noticed that too. And as a second question, with DRAM pricing having gotten notably more elevated, how are you thinking about that with respect to gross margin? Thank you.

Appreciate that, Meta. What I'd say is the scale-across opportunity is starting to take shape, and of course we've put out the 51.2 Tb router, which is going to help us pursue it. Broadly speaking, in Q1, almost everything we saw amounted to a deepening of the use cases we already had. We were hoping to create new ones, but honestly, we didn't get a huge number of new orders from them over the course of the quarter, so none of those have been recognized yet. Our view is that the scale-across opportunity is going to keep expanding and developing. Once again, we experienced a fairly significant acceleration in pluggable optics, which is a genuinely strong business for us and hands us one more opportunity to chase with these customers. As for the DRAM question, Mark, I'll hand that over to you.

Right. I'd simply note that with memory, PCB, and optics alike, we've observed supply getting somewhat tighter. Regarding memory, we've witnessed the same thing all of you have, and that includes fairly substantial price hikes too. Yet when it comes to both availability and cost, each of those factors has been factored into and accounted for within our revised outlook for Q2 and for the full year.

Thank you, Meta. Michelle, can we move to the next analyst?

Hi there. Fantastic quarter. With such a strong quarter, asking a difficult question feels awkward, yet I'll go ahead and do so if that's alright with you. Suppose I take away $1 billion from the prior year's revenues, those being the AI backend, and I also take away $3 million from the guidance for the coming year, then what's left of the business, which makes up the majority, comes to a $55 billion base for last year, and that's expanding at just 3.6%. Why aren't we seeing bigger growth given what we have in Wi-Fi, campus, and security? Why isn't growth above 3.5% showing up for the remaining portion of the business?

Tal, that's a fair question, and asking tough ones is always something you're permitted to do. What I'll do is highlight the orders picture for Q1, and after that I'll hand it to Mark to say a bit about the P&L. Let me be clear: once you normalize out the Hyperscaler growth in Q1, orders for the remaining portion of the business were up 9%. We haven't previously shared that figure with you. I'd like you to have it. Mark, would you like to comment?

Okay, Tal, I appreciate you asking that, Tal. What I'd say is that we're 90 days into the year. We've had a really strong beginning, which you can see in how orders have grown and in the momentum we're building. Once we move into the back half of the year, the comparisons become far harder. In other words, the comps during Q1 and Q2 from a year ago were minus 6% and plus 9%. When you reach Q3 and Q4, those comps are plus 11% and plus 8% on the top line. The comparisons get significantly tougher there, yet I do track what you're calculating, and that seems roughly correct.

Thank you, Tal. Michelle, can we move to the next analyst, please?

Hi folks, how's it going? Chuck, I'd like to ask about something you said regarding how these cycles span multiple years. You've given guidance for the year, and we're fine with that, but could you expand on it a bit more? What are you observing that makes you comfortable discussing multi-year cycles, perhaps call out the main ones you have in mind in a bit more detail, and what prompted that remark, since it stood out as especially interesting? Thanks a lot.

Yes, thanks, Ben. Regarding the refresh opportunity, as you may remember, we introduced a fresh set of products in enterprise routing, we introduced a new Wi-Fi 7 lineup, and we introduced a fresh set of campus switches. What we observed is this: consider the Cat 4K and Cat 6K installed base approaching end of support — that's one element pushing this; as we scaled the launch of these new offerings, all three product families are ramping more quickly than in prior launches. That makes us believe customers truly are moving on this aggressively. Another point is that it shows customers remain quite focused on updating their enterprise network infrastructure to get ready for inferencing and AI workloads. These efforts always stretch across multiple years, Ben. Back in 2017, when we introduced the Catalyst 9K, that shift simply continued for five, six, seven years. That's just what we typically observe with these transitions. The fact that they're ramping more quickly than past experience suggests there's substantial interest in these portfolios. Finally, I'd note that these new switching platforms, which genuinely allow security to be deeply fused into the network, carry a message that resonates with our customers, especially as they consider agentic workflows ahead. They understand they can't send this traffic over to a firewall. As traffic moves, they'll need to apply security policies within the network. I believe that's one more factor they're getting ready for.

Appreciate that, Ben. One more thing — I'd ask the analysts to pose both their question and a follow-up together. Michelle, let's go ahead and bring in the next analyst.

Hey, everyone, great results on the networking front. So where do things stand — what's the current level of Silicon One penetration within the product lineup? Looking at it as a standalone offering, what do you believe is driving Silicon One to pick up that traction among the Hyperscalers, and what advantages does it bring compared with the custom solutions people keep mentioning, plus Broadcom's newest chips? And as a second question, why is it that Splunk is now beginning to experience those bigger migrations toward cloud, and do you have any read on how that affected the security figure this quarter? Thanks, everyone.

Yes, Jim. Regarding Silicon, our expectation is that by the close of fiscal 2029, meaning roughly another 2.5 years out, we will have it completely deployed. Silicon One will be fully deployed throughout the whole portfolio. That's what we're aiming for. It's already inside certain data center switching offerings headed to the enterprise. It's clearly inside the Hyperscaler offerings. My view is it comes down to a mix of performance, programmability, and low power consumption. Another factor, I'd say, is that they like having multiple sources, and they genuinely value the custom engagements we do with them, digging into their specific needs for each one. I believe that's genuinely what has shifted the trajectory for us with these customers. For any of you who joined these calls 5, 6, 7 years back, you'll recall I kept saying we hadn't performed well in this area; we hadn't performed well in this area, and that we intended to change that. I'm genuinely proud of what the teams have achieved. On the security topic, Jim, I'm glad you raised the revenue issue. As I mentioned in my prepared remarks about Splunk, and then Mark, I'll let you speak a bit about what it implies. We simply observed a fairly significant mix shift toward cloud during the quarter. My recollection is the previous quarter was roughly 50/50, and I believe on-prem ended up being just about one-third of revenue. We had anticipated a mix closer to 50/50, since that's what we'd been seeing. Ultimately, it's actually quite positive, as I said, since it lets us drive adoption with our customers, expansion, and deliver innovation in real-time, because it's cloud-tethered. Over the long term, that's good. During the quarter, it was clearly a one-time timing matter on the revenue side. If you examine all the dynamics on the order side, things are broadly as they were in the previous quarter. We saw positive order growth in our new and refreshed products. We saw the same drag from our prior generation products, though that figure keeps getting smaller; we saw double-digit ARR, double-digit RPO on the product side as it relates to Splunk; on the demand side, there's not much to worry about. There was simply that nonrecurring revenue matter tied to the recognition under the 606 accounting standard. Mark, is there anything you'd like to add?

Right. So when you consider the shift we're observing on Splunk's side, like you said, Chuck, that's a positive development. Even with the timing gap you'll notice due to ASC 606 and how revenue gets recognized for on-prem versus cloud, we anticipate recognizing additional revenue as time goes on. Looking at our cloud offerings, they hold onto customers more firmly compared to on-prem options. Customers can pick up the technology more quickly, take on features more quickly, and so on. That's a beneficial thing. What you're seeing in, I believe, a stronger indicator of Splunk's business health is really examining ARR and RPO. For the quarter, both increased at double-digit rates. Once more, though we're a bit let down by the timing during the quarter, on the whole, this is genuinely a favorable situation for us.

Thank you, Jim. Operator, we can move to the next analyst.

Thanks a lot, everyone, for fielding my questions. First off, perhaps Chuck, to kick things off, let's shift over to campus — I believe in both the deck and your prepared comments, you mentioned that next-gen solutions are being adopted more quickly than previous product rollouts. Could you help us understand what's behind that? Is competition a factor? Is it simply that these are superior offerings versus rivals? Are you noticing any disruption coming from smaller competitors — the number two and number three players in the market — given the way things are playing out with that kind of integration? Any additional detail would be useful; and on that same note, perhaps regarding the campus market, I recall you noted during the prepared remarks that plenty of end-of-life product exists out there. How are you thinking about that in connection with those customers — I believe they're government customers, or at least public sector ones — when you factor in what's happening in D.C. at this moment? Is upgrading them as soon as funds become available and start flowing part of the plan moving forward? Just attempting to connect these two markets — campus and government — and what's occurring in each.

Yes. Thanks, David. On the campus side, I'd say it's largely what I mentioned earlier. I believe it's the end-of-sale items — the Cat 4K, Cat 6K, and older Wi-Fi. What we're seeing in these initial days is adoption that's a bit quicker than what we've typically experienced with similar launches. Regarding number two and number three, there's definitely been some marketplace confusion, especially when it comes to Wi-Fi. We've discussed that across multiple quarters now. I think that's worked in our favor. I can't offer any specifics on whether that represents a significant portion of it currently. The other two factors I believe are fueling this are AI readiness and the conviction that security ultimately must reside in the network; we're the only company possessing both security and networking. You've watched some of our rivals announce partnerships with security vendors, and those are difficult to execute. I mean, achieving the degree of integration you'll need is hard when you don't own each of the technologies yourself. We believe we're strongly positioned in that regard. We think every one of those elements is influencing various customers. When it comes to the government, we were happy last quarter that our US federal business, even with the shutdown, saw high single digits growth in terms of orders. We considered that pretty encouraging. We're hopeful that the vote takes place today, that the government reopens, and that we won't need to discuss this any further. When it comes to the whole end-of-support situation, which is especially relevant in the federal government, there's a great deal of conversation and mounting pressure around making sure that hardware gets refreshed, purely from the standpoint of cyber risk and good hygiene. We'd anticipate some benefit from that once they return, start working once more, and the government reopens.

Thank you, David. Michelle, we can move to the next analyst.

Hi. I appreciate you fielding my questions. Chuck, perhaps I'll aim both of them at the AI topic specifically. To start, regarding optics, you mentioned the impressive expansion happening with Acacia. Could you expand on optics more generally — what sort of demand are you observing within the data center as opposed to beyond it, particularly around pull-through from scale across and similar areas? In what ways is Cisco involved in each of those? For my second question, I caught you mention two times the volume of orders regarding AI from that same set of customers in fiscal 2026. With sovereign customers, it appears engagement is moving forward yet orders haven't really materialized so far; that would represent potential upside to that figure. I simply wanted to confirm that sovereigns aren't counted among those orders arriving within that roughly $4 billion figure you're pointing to for orders. Thank you.

Yes, Samik, I appreciate that; when it comes to optics and AI, we're involved in both areas. We're fortunate to offer excellent solutions both within the data center and beyond it, DCI, scale across. I believe we'll keep pushing innovation there. Like I mentioned, we're currently selling our pluggable optics to every one of the Hyperscalers, all the big Hyperscalers. I think we'll keep seeing fantastic chances across both areas. Once more, it's a genuinely massive market that matters a lot, even comparable to the switching portion of it. We're happy to be engaged on both fronts; regarding orders, yes, what I meant was that we anticipate this year we'd reach no less than 2x the orders we received from Hyperscale customers during the prior year. That involves the identical products we tracked last year and the identical customers we tracked last year. Nothing has shifted; when it comes to Neocloud, Sovereign Cloud, Enterprise, there's that potential upside where I noted we had more than $2 billion of pipeline remaining for the rest of the year. Not a single bit of that is part of that 2x figure I referenced. You got that completely correct.

I might just add one more thing, Chuck, regarding the sovereign piece: the initial stages of that sovereign rollout truly don't carry much weight for this year's guidance either.