Cisco Systems, Inc. (NASDAQ:CSCO) Goldman Sachs Communacopia + Technology Conference September 10, 2024 11:10 AM ET

Company Participants

Scott Herren - EVP and CFO

Conference Call Participants

Mike Ng - Goldman Sachs

Mike Ng

Great. Well, thank you everybody. Welcome to the Cisco fireside chat at the Goldman Sachs Communacopia and Technology Conference. I have the privilege of introducing Scott Herren, who is the Executive Vice President and CFO of Cisco. Scott joined Cisco in 2020 from Autodesk, where he served as CFO for six years. My name is Mike Ng, and I cover Cisco and networking and hardware here at Goldman Sachs. We have about 35 minutes for today's presentation.

So first, thank you so much, Scott, for being here with us today.

Scott Herren

Thanks, Mike. This is such a great event every year, and it's nice to have a chance to join it with you.

Question-and-Answer Session

Q - MikeNg

That's very kind of you to say. Maybe we can start with a big-picture question. Cisco, obviously, the leading networking technology company in the world with leading market share positions in nearly every market and product that it operates in. Could you just talk about some of the networking trends that are shaping technology demand today and how you think Cisco is positioned to address some of these trends?

Scott Herren

Sure, sure. Nice to start with an open-ended kind of question. So we are in the beginning of our next fiscal year. Our fiscal year ended at the end of July. So we had our fourth quarter conference call about a month ago now in mid-August. And what we saw is what we expected to see. We saw a return of product demand as customers really worked through and digested the huge amount of product that we had shipped to them in three consecutive quarters, right?

Q3 and Q4 of '23 and actually the first quarter of our fiscal '24, which just ended, we shipped out a tremendous amount of excess backlog as the supply chain issues cleared up. And so it moved from our bottleneck to their bottleneck. And we saw them working that down and we knew once they finished working through that, that what we would see is a return of demand. That's exactly what we saw.

And it was actually quite balanced demand, which was -- which is also a good sign, balanced by geography. So it's our first full quarter of having Splunk. So let me give you the numbers without Splunk included. So we had product order growth in the mid-teens, inclusive of Splunk. But product order growth of 6%, which was 6% in the Americas, 6% in EMEA and 8% in APJC.

And if you slice it the other way, so not by geo, but by customer market, it was good growth in enterprise, double-digit growth in public sector and actually good growth in service provider and cloud for us, which that customer market, the way we group customers together there includes the cable and telco, which have one set of dynamics going on and hyperscaler, which has a very different set of dynamics going on inside there.

So we saw good growth across customer markets, good growth across geo and -- geos and good growth across product lines. You saw, for example, we had double-digit growth in our securities business, which you've been around us, Mike. You know that's a business we've really turned around over the last couple of years and seeing double-digit growth again in the security business was quite encouraging. We actually had double-digit growth in our collaboration business as well during the quarter and good strong growth across many of our networking segments.

So it was good balanced growth. And I think that the other sign that I think is a positive for us is campus networking actually had nice growth during the quarter. I think that's one that there's been a lot of what's going to happen. There's vacant office space, are people really going to refresh their infrastructure? And we saw really nice growth across campus as well.

Mike Ng

No, that's great to hear. And as you mentioned, we went through this period of inventory digestion and you're clearly seeing an inflection in a lot of your products, customer types and regions. Is it too early to say that we've inflected off the bottom and we're back to a more normal level of growth or are there any nuances that you would call out as it relates to whatever the segment may be products, verticals, regions? Yes.

Scott Herren

Yes. No, I do feel like the trough is behind us at this point. And you've tracked and you've seen how product order growth have inflected over the last couple of quarters. So no, I think we're back onto a more -- we're getting back onto a more normalized demand pattern that you'd expect at our scale.

And as I said, it's broad-brush. I think one of the other interesting points inside that growth, I talked about seeing growth in service provider and cloud, which for us is cable and telco, which has one set of dynamics, hyperscaler, which we know are investing heavily right now.

If you just looked at the hyperscaler piece of that, the second half of fiscal '24, we had north of 30% growth in hyperscaler. And I think that's a number that's not obvious. It's not easy for you to discern from the outside looking in, but it's more comparable to some of the people that you see that are more focused on that space. It's a more comparable growth rate.

Mike Ng

It's great. As we kind of start at the onset, Cisco has a very strong market position in nearly every product that it participates in. That's an admirable position to be in, but there's also challenger companies and competitors that seek to gain some of that share. Maybe you can talk a little bit about Cisco's competitive moat, its differentiation, how is it positioning itself to defend and continue to gain market share in some instances?

Scott Herren

Yes. So I'll start by saying innovation is the cornerstone of that. And we're investing at this point a little more than $8 billion a year in R&D driving that level of innovation. You've seen Hypershield come out, which is a really interesting melding of both security and networking. It builds the security into the fabric of the network.

It's a tremendous opportunity. It's kind of the next wave of where we see the kind of not just selling security or selling networking, selling secure networking, right? And I think that's a space that's going to be -- we're going to be extremely well-positioned to do that. No one's got both the depth of security and the depth of networking that we have.

So innovation is kind of the core of that moat. But the level of trust that we've built up, not just as a provider, but also in terms of the service that we provide to our customers is significant at this point. And it's with a massive global footprint, becomes a pretty significant buying or buying criteria for people. We can support them wherever they are, 20,000 plus resellers worldwide. There's a huge amount of just the scale that we have and the footprint that we have that aligns very nicely to many of our customers. That's an advantage.

But I think the third, and we've just recently made some changes to further augment this is the breadth of our product portfolio. That's -- no one has the footprint, no one can match the footprint we have in networking plus security with Splunk in there now, plus the core data infrastructure that you need and with our collaboration tools and being able to bring that all together and leverage the advantage of the breadth of that portfolio instead of trying to sell piece parts individually against piece part providers is a significant opportunity for us and a significant advantage that I think only we have in that space.

And so one of the things that we announced also in the fourth quarter earnings call is Jeetu Patel taking over as our Chief Product Officer and bringing all of our product lines now under a single person and not only is it having it all under one person make it much easier to drive those cross architecture synergies, Jeetu is one of the most talented product people I've ever been around. I mean, he has -- before this was running our collaboration business and I think built by far the best collaboration product out there.

Obviously, it's difficult to sell against the perception of free, but built a great product there, has done a great job within collaboration on devices and with our contact and online or cloud-based calling.

He then took over our security business and has really turned that around both by bringing the right level of talent in at the leadership level and being very innovative in things like Hypershield and building that into the fabric of the network, our firewall line has been almost completely refreshed. The ultra-high end is still to be refreshed. And we've seen really nice results. We've seen a couple of quarters now of double-digit growth in security.

So he's very talented. He's got a very high sense of urgency. And I think having one person where it all comes together to drive the benefit of the breadth of our portfolio is going to be a significant advantage for us.

Mike Ng

Yes. I mean that's a great segue to maybe just talk a little bit more about security. To your point, Scott, Cisco's securities business has seen a lot of success in the last couple of quarters. The growth rates have certainly improved. Maybe you can provide a little bit of background of where we've been and where we are now. You started talking a little bit about that just earlier, but maybe you can just lay the groundwork for why security is doing so well now.

Scott Herren

Yes, I think it's a number of things that are driving that. The first is we had historically, and by that, I mean go back to when I joined Cisco. Our security strategy was really to -- more of a piece part strategy to sell against other piece part providers, which I guess made sense at the time. It was clear to me that, that wasn't playing to the advantage that we just talked about the breadth of our portfolio.

So we made some changes in leadership there, brought in some really talented people to run that from that are security industry experts. And then having a talented guy like Jeetu who is one of the most innovative, creative people I've ever been around kind of drive that has led us onto a much more of a security platform approach. That's why the Splunk acquisition made perfect sense. The SNAP and as the SIM that's kind of the back-end of our security infrastructure on top of that.

We've launched some really innovative new offerings there, our XDR offering there in less than a year has scaled to almost 600 customers there. And so we're seeing that our Secure Service Edge, SSE, offering, which is also new, is also scaling quite nicely and our cross-security portfolio, AI assistant is now being used by more than 2,000 customers around the world.

So it's really going away from trying to be piece part competitive to saying, you know what, let's pull this together into a broader security strategy that our customers can implement. And by the way, open to having third-party products SNAP into that infrastructure, which I think you have to be in security.

Mike Ng

Right. And that's a good place to start talking about Splunk. Splunk naturally, I believe is part of the security and observability segments within Cisco. So could you just recap the strategic rationale for Splunk now that we're five to six months in, provide a mark-to-market on where we are today, things that may have been better than expectations, things that may need a little bit of work?

Scott Herren

Sure. Yes, the Splunk integration, as you said, we're probably in month six at this point on the Splunk integration. If I step back to the thought process behind that, you know we've been building out a much more recurring revenue model at Cisco, which brings with it the advantage of predictability and more than 50%, post-Splunk about 56% of our reported revenues are now on a subscription basis, right?

So as -- we're continuing to build that out. So it made sense from that standpoint, it fit perfectly into our security platform strategy with the SIM back-end as a company on its own, it's extremely sticky, right? It gets deeply embedded into our customers' environments and it's extremely sticky and that's continued post-acquisition. They've continued the momentum that they had when we acquired them.

But it also fits really well from a cultural standpoint, which is really important. Not only did we get good product, we got a cultural fit and some great leadership talent. The CEO, Gary Steele, who is a guy I've known a long time. Gary, before coming to Splunk, was the CEO of a company called Proofpoint. If you know security, you know Proofpoint.

And I served on his Board before I joined Cisco. I served on his Board for four years. I've known Gary a long time. He's a very focused operational exec. He's now President of our Go-to-Market worldwide. So the entire Cisco Go-to-Market engine reports to Gary and he is an extremely talented person to start to drive that.

So we got good leadership, got a very sticky product that's growing very quickly. It's accretive to our gross margins in year one. It will be accretive bottom line in year two, which is our fiscal '26. That's cash-flow accretive outside of the cost of the transaction, cash-flow accretive from the start, even with the cost of servicing the debt or the foregone interest from cash on the balance sheet. So it's a nice fit financially. It's a nice fit culturally and it's a great fit strategically.

Mike Ng

Yes. I didn't cover Splunk or look at Proofpoint, but the investor feedback that I've got on Gary has only been positive. So maybe you can talk about Gary's expanded role as Head of Go-to-Market, like what's the secret sauce that Gary is going to bring to the rest of the organization.

Scott Herren

That's a handful of things. He's a very focused operational, roll-up the sleeves, get into the detail and with a company at our scale, that's not an easy task. He is absolutely tireless. Like I said, I've known him for a long time and I've watched him. He scaled Proofpoint. He's unique in the way he can scale.

When he became the CEO at Proofpoint, they were running around a $15 million annual revenue rate. He scaled that to $1 billion, while he was there over the course of time, quickly turned around Splunk in about two years from a company that was not profitable and that was seeing some stagnation in their growth rates to growing in the mid-teens again and very profitable.

What he brings is besides obviously having great experience as a sales leader, right, which is what he was earlier in his career, he brings a fresh set of eyes to look at the like why are we doing it this way. And it's been absolutely terrific working with him as he looked at things and kind of said, you know what, we could do this differently. We could do this in a better way. And I think that -- it's hard to find people that can come in at that level of scale and still have what I call fresh eyes to look across it and come up with things that need to be done. But he's also a no-nonsense kind of a person. He's very direct, which I love. I tend to be fairly direct as well. And I appreciate that style. He's fitting extremely well with the rest of the executives team.

Mike Ng

And he's got the track record of success to prove it all. Maybe one last one on security. To your point, Cisco is pursuing security with more of a platform-based approach. It feels like that's a little bit different than what's been in the market in terms of best-of-breed, working with like most multiple security point products. And I think that's led to some customer frustration, managing so many different vendors and security. So is it your view we're moving towards a platform world and is Cisco the leading vendor to be able to deliver that?

Scott Herren

Yes, absolutely. I mean, you will never talk to a CIO, I don't think, who will say, I wish I had more vendors to work with, right? What they would all love to see is consolidation. Security is a space where you -- consolidation cannot lead to less efficacy at the same time. And so what we've done is as we've built out that security platform and obviously, Splunk becoming one of the foundational elements of that is really built it with an eye toward more bundled product sales across our security line, but also keeping it open with a very robust layer of APIs around it for third-party product integration into that platform.

But if you think of what drives a more proactive security strategy is data and then the ability to apply AI, various AI techniques to that data. And with the data we already have, the networking telemetry we have from the breadth of our footprint and networking with the huge amount of telemetry we have from our security products from still continuing to have one of the most robust security endpoint products that's out there.

And then with the Splunk log data at the same time. Being able to pull that together and put AI techniques on top of it gives the security team an ability to not only consolidate vendors, but then to very seamlessly say of the 1,000, 5,000, 10,000 security incidents that I have today, what are the five that I have to get on immediately, what are the five that are the most important, that's an AI opportunity and that's what we've already started to build into our security products because of the breadth of the telemetry we have that we can feed into that AI engine.

There is an ability to do dialogue with a security ops center. You've been in a security ops center, like a wall of screens and each one has a different -- it's like command control for the lunar launch. And you look at that and go like how do I know really where I need to focus in. And so there's a huge opportunity.

But not only to say here's the most important to say, hey, with this incident, here's what I see, here's what I recommend is the next step, would you like me to take it? And the security ops person can say, yes, and the system will often take that step and it will say, here's the next thing we could do. If it's an incident on an endpoint, do you want me to isolate it from the network? We pretty much do want it to isolate from that.

Do you want me to dispatch a new device to that person whose endpoint is so they can continue to get their job done? Probably a good idea. Do you want me to run forensics now on that device and see where -- was there any exfiltration or was there any movement laterally across the network? Yes, I want you to do that as well.

So you can build this kind of dialog capability. We've already demoed some of that you can build into it. So it's a -- I think it's a pretty significant opportunity for us. There's not many that have the breadth that we have to be able to build it out with that level of capability.

Mike Ng

Great. One product category I wanted to touch on also is wireless LAN. Obviously, there's a lot happening in the space from some of the M&A that's in process right now, but also AI and cloud innovation to support wireless LAN. Could you talk a little bit about Cisco's positioning here? Are you seeing opportunities because of HPE and Juniper?

Scott Herren

I think for sure that's created just a degree of uncertainty and a question of, hey, should I consider if I was previously a vendor or a customer of either of those, now is the time to kind of open up and look at other opportunities. And we've seen our wireless business, our orders greater than $1 million grew more than 20% in the fourth quarter.

So we're seeing scale in that wireless space. We're also pulling together our own -- our Catalyst and Meraki lines under a single cloud platform, a back-end plane to drive that. So we're seeing pretty significant opportunities there in the wireless space. Some of it could be generated by the M&A that's in the background. I think that's opened a lot of people's eyes. It's also generated by some of the innovation that we've built into that space.

Mike Ng

Great. I wanted to ask about some of the financial guidance points before talking about AI, but it's encouraging to hear Cisco's fiscal '25 revenue guidance of $55.5 billion to $56.2 billion, which is about 3% to 4% revenue growth. Could you talk about some of the trends that you're seeing in Cisco's business and the industry at large that give you confidence in this fiscal '25 outlook?

Scott Herren

Yes. And it's 55% to 56%...

Mike Ng

Sorry.

Scott Herren

…top line. That's okay. So yes, we do see a number of dynamics. We talked about that, obviously, you need to see order growth ahead of seeing revenue growth. And for us, by the way, the statistic that's not part of that guide, but that you see in our results is our RPO. A lot of those product sales for us are subscription-based products that first have to go through deferred revenue before they accrete to the revenue stream.

We built up right at $40 billion of RPO at this point. Roughly half of that will accrete to revenue in the next 12 months, all right. So there's not just the revenue that you see, there's the revenue that's built up that will be future revenues for us sitting in the RPO balance.

There's a couple of factors inside there. Obviously, adding Splunk to the portfolio, there's a degree of inorganic. We, in fiscal '24, had Splunk on board for about 4.5 quarters, so a little more than a third of the year. So there's going to be some inorganic growth in there driven by Splunk.

The compare point, if you remember, our first quarter of fiscal '24 was actually very strong and it was really the last quarter where we shipped out a fair amount of excess backlog that we have built up because of the supply constraints. So that becomes a tough compare point for us in this current -- we're in our first quarter of fiscal '25, becomes a bit of a tough compare point for us.

And that number was not insignificant. The amount that we shipped out in Q1 of '24 was in the $4 billion range. So those two kind of counteract each other. I think as we look ahead, seeing that the trend that we've seen in product orders and the inventory digestion behind us now and seeing it be balanced across geos and across customer sets, it's fairly encouraging. I think it's also a time to be prudent. The economic calendar, the economic -- the world economic events, elections around the world, it's a good time for us to be prudent. I think you'll see some of that built into our guide as well.

Mike Ng

Yes. And the trajectory seems interesting where we'll have one more quarter of tough comps because of backlog…

Scott Herren

That's right. The quarter that we're in.

Mike Ng

…ago and then we'll see revenue growth accelerate. Okay. At Cisco Live earlier this year, Cisco outlined fiscal '26 and fiscal '27 revenue growth targets of 4% to 6% and you gave a lot of detail at the segments as well. Could you talk about why 4% to 6% was the right growth outlook for those two years?

Scott Herren

Sure. Yes, well, within that product is growing 5% to 7%. And then we gave you a little bit of a breakdown what was happening within network, inside that, what was happening in security and observability and affect collab. There's a few other things in there, but affect collab. And obviously, you see significantly faster growth within security and observability. Networking business will continue to grow, but there's such a blend with the breadth of our portfolio, there's such a blend of different products inside there.

And I think that's -- again, it's easy to get confused when you try to compare our networking growth rate to some of our peers who are kind of focused on a single segment. And that's why I think it's important that I -- the stat that I gave you earlier, which is if you just break down that networking growth rate over the last six months, our growth in the hyperscaler space has been a little north of 30% over those last six months of the year.

So seeing really nice growth in that space, but there's a whole blend of products inside there and it's a very large number for us sitting in that space. So I think the other thing that factored into that guide, obviously, fiscal '24 was a tough year for us. There was a lot of change that we didn't see coming necessarily. So it pays to be a little bit prudent at this point as I'm giving projections and there's a little bit of that built in as well.

Mike Ng

Great. And just to follow up on that hyperscale point, maybe we can talk about AI. During last quarter's earnings, Cisco talked about achieving $1 billion of AI orders from webscalers to date and they -- and you reiterated the additional $1 billion of AI orders in fiscal '25. Maybe you could just clarify like what are you doing with AI orders? Is it primarily back-end networking? Is it mostly webscale? Yes.

Scott Herren

Yes. So those numbers are almost exclusively webscale. There's very little -- historically, it's all webscale. As we look ahead, it's almost -- the overwhelming majority is also webscale there. Historically, it's been kind of three things that drove the $1 billion of orders and those are inception to date. Those weren't just in fiscal '24.

Mike Ng

Right.

Scott Herren

It's the top-of-rack, the 8k for us, the 8000. It's the optics and optical are a piece of that. And Silicon One is a small piece of that. There are some of the large webscalers that just want to buy the Silicon One chipset from us. As we look ahead, it's going to be much more. So that's what's in the $1 billion inception to date.

As we look ahead, and again, it's webscale focused, it's going to be a little bit more heavier on the system side, on the 8k side with also some optics built into that. You know, that when you sell AI capabilities into these large hyperscalers, it's a design win, right? They each have slightly different designs that affect both the underlying chipset, but of course, how you build that into a system.

And so what we've been in the process of, it's an engineering to engineering process, working with them on understanding what their needs are exactly, designing that product for them, going through the process of having it certified and then you've got a design win. And then what they look to do is buy off of that design win.

So we've got multiple -- we gave some of the stats at our Investor Day design wins in that space. That's what fuels the $1 billion that we have line of sight to the next $1 billion that we have line of sight to in fiscal '25 is understanding exactly what those design wins are.

What's not in there is what we see selling into the enterprise, one of the interesting slides that we had on our last call laid out just four of the large transactions that we got done during the quarter, each of them greater than $100 million sales.

And when you looked at what they were buying inside there, it was the things that people always think of buying from Cisco, switches and routers and WiFi and security and collaboration. But when you looked at why they bought that, it's really refreshing -- in the enterprise, refreshing their core infrastructure to be ready for AI.

When they are ready to actually start implementing AI at scale, they all know that they need to refresh that infrastructure and that's one of the things that's driving longer-term demand. That is not part of the $1 billion that we talk about of AI. Think of that as hyperscaler. That's a significant opportunity that drives the core business over the next several years.

Mike Ng

Yes. There's a pull-along effect that people make AI infrastructure investments. Just on the hyperscaler piece, I mean, Cisco has been very candid that it's a customer segment that they had not done much with in the last several years. And this is a notable change, right, and improving of that customer relationship. Maybe you can speak to a little bit about that and where we are in the hyperscaler opportunity.

Scott Herren

Right. Right, right. Yes. If you look historically at the hyperscaler space, go back a decade-plus when the big compute and storage capabilities are being set up, we didn't play very much in that space. And part of it was we didn't really recognize this whole design. It's a design win. You have to design separately. Chuck came in as the CEO and quickly said, look, this is a space that we need to be more effective in. That's when we started the Silicon One strategy, the Silicon One chipset strategy.

But it's when we also went back to the hyperscalers and said, look, we'll meet you where you are. If you want to buy completed systems from us, we'll sell you that. If you want to buy a white box from us, we'll sell you that and put your own software on it.

If you want to buy just the chipset from you, we'll sell you that. And what that led to is us even before the AI infrastructure build-out, building a pretty robust business in that hyperscaler space. So clearly, we gained a fair amount of -- from zero, it's easy to gain share when you're zero, right? We gained a fair amount of share in that space.

Now as the AI infrastructure build is being built out in that same set of customers, we've got relationships with them. We've got the you'll meet you where you are kind of a mindset. We understand how to build out those design wins. We recorded around numbers 20 of those already.

So we're in a -- we're much better positioned for the build-out of the AI infrastructure there. As I said, second half of last year, that business grew a little bit north of 30% for us. So we -- I think we're very well-positioned as that builds out in the future.

Mike Ng

Great. And could you talk about AI infrastructure demand among enterprise customers? Cisco has a partnership with NVIDIA to help service some of these customers. I think there's a little bit of a debate in the market, whether or not that enterprise AI infrastructure demand will materialize and win. Obviously, Cisco is going to be in a very good position to service those customers because of the channel footprint, the sales footprint and everything like that.

Scott Herren

That's right. Yes, it's a partnership that I would say is still in the very early stages. We've spent an enormous amount of time working with the NVIDIA team. And that's the -- when we recently launched this HyperFabric capability. Here's what that is. Think of that as if you're an enterprise and you want to train a model, you're likely because of the scale of training, you're likely to do that in the public cloud.

But when you want to run the inferencing on a regular basis, because of the sensitivity of the data both in and out of those models, the enterprises are largely we hear this pretty frequently from our customers, they want to build that out within their own data center, move away from the public cloud for the inferencing phase. That's the opportunity that HyperFabric will go after.

And what that is, is inferencing in a box, right? It's a single purchase that will -- it will be a rack, it will have servers inside there with NVIDIA GPUs on it, our control plane on the back-end and our top-of-rack Ethernet on the top-end of that. But it's the easy way for enterprises to kind of within one purchase get a single rack, the HyperFabric rack that will support their own internal inferencing needs. And we hear repeatedly that that's -- while that market is somewhat nascent right now and what we see from enterprises is they're building out the rest of their infrastructure to support AI, that opportunity for HyperFabric is going to be fairly significant over the next few years.

Mike Ng

Okay, great. Let's talk about capital allocation. Cisco returned, I think, over $12 billion of cash to investors last year through repurchases and dividends, more than 100% of free cash flow. What is Cisco's shareholder capital return and other capital allocation strategies? The Company has historically operated at a net cash position. You've got some net debt because of Splunk. Does that kind of preclude you from doing any additional M&A at this point?

Scott Herren

Yes, not at all. So we did move from a net cash position through the Splunk acquisition to a net-debt position. Our ratings, by the way, our balance sheet is very robust, as I'm sure you know, Mike. So AA minus A1 rated, didn't change through the course of that.

We -- at our scale and with our margins, we generate a lot of cash each year. So the cap allocation strategy has not changed. And just to recap it, support the growth of the business has got to be job one, but job one-A is to continue to support the dividend and grow the dividend. You see us grow that every year, grow that dividend. Right now, that consumes about $6.5 billion of cash per year.

We've also, for the last several quarters, said our target is $1.25 billion per quarter, so $5 billion a year of share buybacks, which is slightly more than offsetting, it's actually materially more than offsetting the dilution of our equity plans. And on top of that with an opportunistic overlay. And you just saw in the last quarter and the fourth quarter, there was some market dislocation on our price relative to the rest of the market.

There's a number of factors that go into that, but we actually instead of just doing $1.25 billion in the quarter we just closed, we did about $2 billion of share buybacks. And so there will always be an opportunistic overlay when we see the number of factors that lead us to believe that there's a dislocation on our share price.

So our cap allocation strategy is unchanged. Our robust balance sheet has the same ratings that it did before. There's no change in the -- it's not on the watchlist or anything like that. We continue to generate a lot of cash and continue to use it to both return capital to shareholders, but also drive the growth of our business.

Mike Ng

Great. In the last minute that we have, maybe you can talk a little bit about your vision for Cisco over the next three years, but could you also just touch on efficiencies, cost savings, which has been something that you guys have done a good job of?

Scott Herren

Yes. I'd say part of what we announced in the fourth quarter, I'll try and do this in the 30 seconds that we have left.

Mike Ng

We can go a little bit over.

Scott Herren

Not going to be easy. Is you've seen us be very disciplined financially in the way that we both invest in the business, but also return good margins across the Board, expect that to continue. The most recent earnings call where we announced the restructuring was really about pivoting more of the resources with our OpEx run rate at this point is somewhere between $19 billion and $20 billion a year. We're spending enough money. What we need to do is make sure that we're prioritizing the fastest growth areas that we serve. And that's AI, of course, that's cybersecurity across the Board and that's cloud and networking.

So that -- what that was about is finding efficiencies across the Company and doing it in a very, I'd say, single tough-minded way of finding those efficiencies so that we can pull those efficiencies out and then staying within our spending envelope, be able to invest more in the fastest growth areas that we have.

So I continue to see us doing that, continuing to be disciplined on pricing, continuing to be disciplined across the Board financially so that we can drive the right level of return, but also drive growth at the same time.

Mike Ng

Wonderful. That's a great place to cap it off. Scott Herren, everybody.

Scott Herren

Thank you.