Hewlett Packard Enterprise Company (HPE) Earnings Call Transcript & Summary

Got it. Really appreciate that detail. And diving a little bit more into edge specific topics here, I'd love to jump a little bit deeper into the Silver Peak comment that you made at the end. It's been a couple of months since Silver Peak was first announced. And so you've had a little more time perhaps even to dive deeper into the details. How do you feel about the proposed Silver Peak deal? And what are the most important aspects of the deal maybe incrementally that you learned in the last couple of months?

Yes. Certainly. So you're right, we have been spending a lot of time, myself, in particular, focusing on that. And we are very excited about the deal and from a few different points of view, right, so one, accelerating HPE's overall strategy, the edge to the cloud platform as a Service strategy, increasing our overall TAM and certainly driving financial synergies as well. But maybe just to take a little bit of a step back for all of those on the line here that aren't familiar with Silver Peak, it's a widely recognized global leader in SD-WAN infrastructure. And as enterprises navigate and plan for a post-COVID world, we expect them to put an increased emphasis on a modern dynamic edge-to-cloud architecture, right? And Silver Peak's SD-WAN platform dramatically improves application performance and also lowers connectivity costs, which is going to be critical to that modern dynamic experience. So just to give you a big picture view about how this deal accelerates our strategy and increases our TAM, first of all, SD-WAN is a $2 billion-plus market, growing dramatically faster than overall IT spending and, certainly, a natural complement and extension of our -- the Aruba SD-Branch platform. So we really believe that we're combining 2 industry leaders in their respective markets, within networking, with this acquisition. And second, we do expect customers to see increased performance and value by bringing both platforms together, especially by integrating Silver Peak along with Aruba's leading ESP platform, which we announced recently, which really uses the power of AI to predict and resolve problems at the network edge before they happen. So that, all in all, we really believe will help to simplify connectivity and that experience across the campus, branch data center and remote worker environments, which are becoming more and more relevant in today's landscape. And then maybe just to touch on quickly here, for the financial synergies, we really believe that those are quite compelling, too. Back in July, when we announced the deal, in addition to driving top line growth, we also stated that we expect Silver Peak to be accretive to our non-GAAP gross margin and neutral to our non-GAAP operating margin by FY '22. So it's really going to help to drive the growth side of the business that Tarek was talking about earlier but in a profitable way as well. So right now, as you had mentioned, we're really just focused on getting prepared for execution. And we really see a lot of the big benefits coming from expanding on Silver Peak's go to market, especially since their business has been very North America-centric to date, and so we expect to leverage our international reach. And they've also had a high reliance on direct sales. So we really plan to scale and enhance the channel partner program and service provider businesses, leveraging those relationships that we already have in place. So just to close on the point here, we continue to expect the deal to close here in our Q4 and for Silver Peak CEO and Founder, David Hughes, to join HPE and to lead our overall combined branch business within Aruba. So just like Tarek was saying before, we'll talk more about this at our upcoming security analyst virtual event on October 15.

Appreciate all that. It sounds -- it definitely sounds like there's a plan that's going to be executed on going forward in regards to the deal integration. I want to switch gears and talk a little bit about Compute for a second. It's a very large segment. It's almost half of your revenue. And it seems like it's on track to be down year-on-year in 2020 given the market dynamics but potentially better than what occurred in 2019 on a calendar year basis. Compute had a solid quarter in the latest quarter. I think it was -- there was a 20-point change in year-on-year revenue declines, nearly flat year-on-year. And admittedly, there were some of these backlog -- the backlog catch-up issues that were resolved in the quarter. But on a constant currency basis, you did grew -- you did grow 1%. But if we back that backlog dynamic out a few hundred million, and that's purely my estimate, not something that you guys have officially commented on, it's still the case that revenues declined potentially nearly 10% year-on-year. Do you expect further stabilization, I guess, in the segment going forward? And what would be some drivers of stabilization for investors to keep in mind?

Yes, sure. So Jeriel, you're right. During Q3, HPE converted $500 million of excess backlog, most of which was Compute-related. And this is why our Compute revenue grew 29% sequentially. So the way I look at the backlog issue is it's simply a deferral of revenue that should have occurred in prior quarters that did not occur because of the disruption of COVID and that we were able to capitalize on in Q3. And so at the end of the day, and if you recall what we said in Q2, we exited Q2 with $750 million of incremental backlog over the standard level, at 2x, the level. So $1.5 billion total backlog, $750 million incremental over the normal level. And that $750 million is revenue that did not vanish. It could have. We could have had some orders canceled. We didn't see any orders canceled. And that's a really good thing, which means that the actual demand was there, and the backlog that we realized in Q2 is a function of orders that were booked in Q1 and Q2 -- excuse me, the backlog we realized in Q3 is a function of the orders that we booked in Q1 and Q2 that we couldn't deliver because the world was disrupted by the pandemic. But the demand is there, and we still continue to expect progress in Q4 on the remaining backlog, approximately $250 million, to get back to normal by the end of this fiscal year. And so that's what has happened. And most -- I have to say most sell-side tried to overanalyze the backlog. It's a fairly simple deferral of revenue within the fiscal year '20 year, and that's all there is to say about it. Normalizing to try to infer a trend of demand is probably more complicated than anything else given that some orders that form part of the Q3 results were booked in Q1, so it's very hard for you, from the outside, to actually determine how much of the backlog needs to go back into Q1 and Q2 without knowing what the order composition is and its timing. So the simple way to look at it is look at fiscal year '20. When we will post our Q4 results, you'll have a complete view. If you were to strip out of Q3 the $500 million and you strip out of Q4 whatever progress we make on the backlog, you will determine what the underlying performance would have been ex backlog, and you will draw your own conclusions. With respect to the other trends that underpin our Compute business, our AUP grew 3% quarter-on-quarter. And we also had a very strong double-digit growth in units quarter-on-quarter, 36%. And clearly, that unit growth was driven by the backlog conversion, but the AUP growth reflects the longer secular trend -- the longer-term secular trend of richer configurations from CPUs and memory. And that growth in AUP was amplified by pricing actions we've taken during the course of Q2 and Q3 to reflect higher commodity costs. Now moving forward, looking to Q4, Compute units and revenues should also see a benefit from converting the remaining backlog, as we said, obviously, in a lower amount than the $500 million. But there is also a secular driver of richer AUP configurations, and we do see that there is potential opportunities for gross margin expansion as we start to see some price pressure or deflation in the cost of commodities moving forward. If you also were to look at our performance in Compute compared to other players with whom we compete against all the way on to Q3, you will -- and you can refer to publications made by IDC and other players, you could observe that we have taken share in this quite difficult market relative to the competition. And that's a really good thing from our perspective.

Got it. We'll see how this continues to play out as you guys continue to capitalize on the backlog and normalize the business. I'd love to jump into GreenLake, which I'm sure you'll cover pretty extensively at your SAM event in a month. And my question is with regards to the ARR metric that you guys had disclosed, which has grown to low double-digit range for the last 3 quarters year-on-year. But the growth has slowed with every single quarter. But at the same time, you've talked about GreenLake orders, and that has been strong, with the recent quarter up 82% year-on-year, so a pretty big dichotomy between revenue growth or ARR growth and orders. Could you give us a sense of how we can reconcile these metrics to understand the forecast for a 30% to 40% CAGR and how to reconcile that versus the slowing ARR metric at the same time?

Sure. So the slowing of the ARR metric is, obviously, if you look at it this way, it's on a year-over-year view and it's just a simple function of tough compares relative to the prior year. The 11% year-over-year growth was in line with our expectations. But most importantly, our ARR has moved north of $500 million. It's been 2 quarters now. And we are at $528 million of ARR. Now when you look at the order growth, what's happening is that we are seeing larger and larger deals. We did announce a large deal in Q3. We could make that deal public, but there are a number of other deals that we could not make public because customers preferred to retain confidentiality. But what we are seeing is that these deals are substantially bigger every quarter. And so we will see, for a period of time, ARR lagging the order growth but then the RR will pick up because the order growth feeds the balance sheet, our balance sheet, and then, obviously, the revenue and the ARR gets recognized as services are rendered. And so it's really important now to look at the order growth and how that order growth feeds the ARR over time. And when you look at that 82% growth, there was an outstanding performance in North America, which delivered a 5x year-over-year growth with GreenLake. Market acceptance varies by geos, and we feel that certain geographies, such as North America, Japan, are within our footprint leading the charge. In other markets, it's -- there is an element of seeding the market and making the market. We have to, in some cases, convince customers of the benefits of an as a Service shift. Most of them are very, very receptive to the conversation. And that's why we feel very comfortable around our ARR growth guidance of 30% to 40% compounded annual growth rate between fiscal year '19 and fiscal year '22. At SAM, we will discuss more GreenLake, as you pointed out, Jeriel. And one of the things we will explain is what makes our ARR in detail and what's truly happening in the marketplace around as a Service on-prem, that it is a buoyant market. Not many analysts see what's going on. As a Service is not only a cloud provider market, that is a very buoyant on-prem as a Service market, and we'll shed some light around this. We feel we have a lead there, particularly due to some of the more traditional players who, in the past week, announced that they want to pivot their business model to as a Service. We'll also share some specific differences around our ARR versus other players who have done these similar pivots. Please do remember that in our ARR, we account only for the portion of the hardware that is sold as an operating lease. For contracts that are typically of a longer duration and more customized, the hardware pull-through from GreenLake is substantial, and this is accounted for as a capital lease with the revenue of the hardware recognized upfront. So we are in a situation where our ARR is surprisingly lower than the true revenues that GreenLake generates as a result of the differences in the contracts and contract durations. I will explain that in much more detail when we will be together at SAM on October 15. But we feel very, very good about the progress we're making there and we are certainly, from an execution standpoint, doubling down on the allocation of resources to push GreenLake and as a Service because, like I said, we see a very buoyant market for as a Service as on-prem.

Got it. I think -- it seems like we have a lot to look forward in about a month regarding GreenLake and understanding some of these underlying as a Service and ARR trends. I guess the last question, I just want to understand, and help investors understand as well, the gross margin dynamics. It took 160 basis point step-down in the quarter. It seems like this is somewhat mix-driven. Compute's a little bit of a drag on gross margins, and it grew, as you highlighted earlier, in our segment discussion, pretty materially quarter-on-quarter. But even if I kind of break down the segment numbers like-for-like, break down the gross margin by segment, which I do in my modeling, it seems it suggests that some segments observe gross margin compression quarter-on-quarter like-for-like as well. What were the drivers of this? And what is implied in guidance?

So our total gross margin at the company level was 30.4%. And it did decline 160 basis points quarter-on-quarter. And that decline is essentially a mix-related effect. As we were clearing the Compute backlog, we had more Compute revenue across the revenue mix. And our Compute revenue is, by nature, the lowest gross margin portfolio, hence, the impact that you have seen. And that was particularly acute as a result of the backlog conversion, which has amplified that mix effect that I just referred to a moment ago. But if you really think of what's happening also specifically within the Compute swim lane and other segments, we are taking pricing actions to offset higher commodity costs which -- and we can see in new orders that they're coming in at higher gross margins. So we are cautiously optimistic that we can capture the benefits of lower commodity costs moving forward. And that will also be beneficial to our inventory levels, and this obviously translates into a lower cost of sales moving forward. With respect to the Intelligent Edge, we mentioned on the Q3 call that margin was pressurized on a sequential basis for different reasons. And this had nothing to do with pricing of actual products or services but more to do with higher logistics and duties costs. As you can expect, during the course of COVID, we had to rearrange a number of -- facets of our distribution business, which translated into higher logistics and duties costs, particularly for Aruba, the edge. But on the whole, we feel that moving forward, we continue to say that we could see some uptick in gross margin in Q4 as a result of the measures we've taken on pricing and some of the effects, such as logistics and duties costs impacts, subsiding moving forward.

All right. Thank you for that detail. I guess it will be an interesting metric to track as the segments unfold -- or as the quarter unfolds. I'll reiterate that HPE does have a security analyst meeting a month from today that will be webcasted. Otherwise, Tarek and Jon, thank you so much for attending our conference. And with that, we'll wrap it here. Thank you so much.

Thank you very much for having us, Jeriel.

Thank you.

Thanks.