Good day, ladies and gentlemen. Welcome to the Intel Corporation second quarter 2016 earnings conference call. At this time, all participants are in a listen-only mode. Later, we will conduct a question-and-answer session, and instructions will be given at that time. If anyone require operator assistance, please press star, then zero on your touch tone telephone. As a reminder, this conference call may be recorded. I would now turn the conference over to Mark Henninger. You may begin.
Thank you, and welcome everyone to Intel's second quarter 2016 earnings conference call. By now, you should have received a copy of our earnings release and the CFO commentary that goes along with it. If you've not received both documents, they're available on our investor website, intc.com. I'm joined today by Brian Krzanich, our CEO, and Stacy Smith, our chief financial officer. In a moment, we'll hear brief remarks from both of them, followed by Q&A. Before we begin, let me remind everyone that today's discussion contains forward-looking statements based on the environment as we currently see it, and as such, does include risks and uncertainties. Please refer to our press release for more information on the specific risk factors that could cause actual results to differ materially.
Also, if during this call we use any non-GAAP financial measures or references, we'll post the appropriate GAAP financial reconciliation to our website, intc.com. With that, let me hand it over to Brian.
Thanks, Mark. Our top-line results for the quarter came in right in line with outlook, and profitability this quarter exceeded our expectations. Year-over-year growth this quarter was 3% overall as we transform Intel into a company that powers the cloud and billions of smart, connected devices. We continue to focus on growth in line with this transformation, as evidenced by results in the data center, IoT, and programmable solutions business this quarter. I'd like to take a few minutes to walk through these results and their implications. I'll start with the Client Computing Group, where we saw a 3% decline in revenue year-over-year this quarter, while operating margin was up 19%. These results were a little better than we expected as the PC supply chain reduced inventories at a slightly slower rate, while the two-in-one and enthusiast product lines continued to grow.
We also started shipping our seventh-generation Core microprocessor, formerly known as Kaby Lake, and our latest LTE modem, known as 7360. Next, the data center, where revenue grew 5% year-over-year. Cloud service providers grew 9%, comm service providers grew 10%, and enterprise was down 1%. We achieved some critical milestones in the quarter that give us confidence in our growing momentum as we enter the second half of the year. In the data center, we're seeing an ongoing preference for performance up and down the pricing stack. Average selling prices increased year-over-year in every microprocessor product segment, from Atom and Xeon D SoCs at the low end, up through Xeon and Xeon Phi at the high end. We continued to gain share in network infrastructure throughout the entire segment as Intel architecture becomes the solution of choice for the transformation of the network to SDN, NFV, and 5G.
The significant share gains at the low end of the network infrastructure segment resulted in an overall 1% decline in data center CPU average selling prices. Progress in the data center extended beyond our CPU product lines. Our latest Xeon Phi accelerator, formerly known as Knights Landing, continued to ramp after shipping the first limited production unit in December of last year. Xeon Phi revenue grew 8x in the first six months of this year versus all of 2015, gaining share in the supercomputing and machine learning segments. Omni-Path, our high-performance computing fabric, was launched earlier this year and has already achieved 30% market segment share of the 100G fabric market. In June's TOP500 supercomputing list, Omni-Path was deployed in half of the new 100G systems, pointing to the performance that this technology brings to the market.
This quarter, we also shipped our first silicon photonics products for revenue, the industry's only fully integrated solution. We expect DCG's adjacent product lines, including Omni-Path, silicon photonics, and Ethernet, to collectively grow more than 20% for the full year, and this quarter, make up 12% of DCG's revenue. The Internet of Things business was up 2% over last year, coming in below our expectations. We saw growth in the industrial and video verticals offset by an inventory burn after a very strong first quarter. We continue to see tremendous potential in this business. A great example was demonstrated earlier this month when we announced our autonomous driving collaboration with BMW and Mobileye, marking a significant step for the auto industry as we work together to establish an industry-standard open platform for autonomous driving.
In addition, we are bringing in-cabin computing technology to power the next generation of BMW's highly autonomous and fully autonomous products, from the door locks to the data center. Our memory business was down 20% over last year and fell short of our expectations as a result of a more competitive pricing environment. While we acknowledge the cyclical and competitive nature of this business, we remain confident in our long-term growth prospects as a result of the new technologies we are bringing to this market. Fab 68 in Dalian, China, started its initial 3D NAND wafers late in the second quarter, but ahead of schedule. We also remain on track to ship 3D XPoint SSDs branded Optane by the end of the year and look forward to delivering this exciting new breakthrough in memory to the industry. The Programmable Solutions Group, formerly known as Altera, delivered great results.
PSG grew 12% over Altera's results last year on strength in comms, infrastructure, and the channel. PSG is on track to ship 14 nanometer Stratix 10 samples this year. I'm very pleased with both the integration of this business and their strong execution. Our security business was up 10% as the restructuring the team completed last year, and their focused execution continues to deliver results. Finally, our restructuring initiative that we began last quarter is solidly on track. This program is changing where and how we invest in everything from research and development to sales and marketing. In April, we announced some important changes to our roadmaps in areas like SoCs and perceptual computing. These changes are accelerating our transformation to a company that powers the cloud and the billions of smart connected computing devices while increasing the profitability in our client business.
In total, we expect this initiative will drive in a net run rate OPEX savings of $1.4 billion by mid-2017. Looking ahead, I'm very excited about the growing momentum heading into the second half of the year. While we remain cautious about the PC segment and continue to expect a decline in the high single digits of this year, we're expecting our businesses outside of CCG to collectively deliver double-digit growth in the third quarter. We are seeing clear signs that our strategy is working, laying a solid foundation for growth built on the data center and the Internet of Things business, reinforced by the combination of memory and FPGAs, and bound together by connectivity. With that, let me turn it over to Stacy.
Thanks, Brian. In the second quarter, we met our financial commitments and made good progress towards our restructuring goals. Our forecast reflects growing momentum as we enter the back half of 2016. Revenue for the second quarter was $13.5 billion, in line with expectations and up 3% year-over-year. Gross margin for the quarter of 62% was approximately a point higher than our expectations, primarily driven by lower platform unit costs. Spending on R&D and MG&A was $5.2 billion, in line with our expectations. We are on track to the restructuring announced on the last earnings call with a reduction of about 6,000 employees in the second quarter. Operating income of $3.2 billion was down 2% from a year ago. The effective tax rate for the quarter was 20%. Earnings per share of $0.59 was down $0.03 from a year ago.
The Client Computing Group had revenue of $7.3 billion, down 3% year-over-year. Client Computing Group operating profit was $1.9 billion, up 19% from a year ago. This improvement is driven by lower overall spending and margin improvements in our mobile products and higher ASPs in the PC segment. The worldwide PC supply chain inventory levels came down a bit in the second quarter, and as we enter the second half, they are at healthy levels. Data center revenue was $4 billion, up 5% year-over-year. The Data Center Group had operating profit of $1.8 billion, down 4% year-over-year, primarily driven by increased costs as we ramp 14 nanometer data center products. As we enter the second half, we expect the enterprise segment of the business to stabilize and the cloud segment growth rate to accelerate.
In addition, we expect increasing ASPs as we ramp our Broadwell-based server products. Our Internet of Things segment achieved revenue of $572 million with year-over-year growth of 2%. Our security business had revenue of $537 million, up 10% year-over-year. Our memory business had revenue of $554 million, down 20% year-over-year. This segment had an operating loss of $224 million as a result of continued pricing pressures, higher startup costs as we ramp 3D NAND in our China factory, and increased 3D XPoint spending. The Programmable Solutions Group had revenue of $465 million, up 12% year-over-year when compared to Altera's results from a year ago. Operating profit was negative $62 million. This includes about $160 million in non-cash charges for inventory adjustments. Excluding these charges would result in about $100 million in positive operating profit.
Total cash balance at the end of the quarter was roughly $17.7 billion, up $2.6 billion from the first quarter. Our total debt is $28.6 billion. Our net cash balance, total cash less debt, and inclusive of our other longer-term investments is negative $5.7 billion. We are projecting to improve this net cash balance over the second half of the year. We are generating healthy levels of free cash flow, which enable us to invest in our business and return cash to shareholders. This is demonstrated in our Q2 results as we generated $3.8 billion of cash from operations in the second quarter, purchased $2.3 billion in capital assets, and repurchased approximately $800 million of stock. In the second quarter, we also paid $1.2 billion in dividends, and as of yesterday's close of market, our dividend yield was about 3%.
As we look forward to the third quarter of 2016, we are forecasting the midpoint of the revenue range at $14.9 billion. This forecast is at the high end of the average seasonal increase for the third quarter. We are forecasting the midpoint of the gross margin range to be 62%. Turning to the full year 2016, we expect revenue growth in the mid-single digits. We continue to expect the overall PC market to be down in the high single digits, and we expect to achieve low double-digit growth in our data center business. Gross margin for the full year 2016 is expected to be 62%, consistent with our prior outlook. You can see our strategy playing out in our first half results and our expectations for the second half.
We expect above-seasonal growth in the back half of the year, led by strong growth in the data center, Internet of Things, and memory businesses. For the year, we expect that growth in those businesses will offset the PC market decline, and with the addition of the Programmable Solutions Group, will result in mid-single-digit revenue growth. Additionally, we are executing to our restructuring program, which allows us to increase investments in strategically important areas, generate financial returns for our owners, and build the foundation for future financial growth. With that, let me turn it back over to Mark.
Okay. Thank you, Brian and Stacy. Moving on now to the Q&A. As is our normal practice, we would ask each participant to ask one question and one follow-up if you have one. Operator, please go ahead and introduce our first questioner.
Ladies and gentlemen, if you have a question at this time, please press star and then the one key on your touch-tone telephone. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. Our first question comes from the line of Christopher Danely of Citigroup. Your line is now open.
Hey, thanks guys. Just a question on the server expectations for the second half. What's giving you the confidence that things are going to bounce back so nicely?
Sure. I'll start, and then Stacy can add. This is Brian, Chris. I think it's really a mix of a couple of things. One, we saw in the second quarter and we project out into the second half, a bit of stabilization of the enterprise side of the market. The enterprise side was down only about 1%, which is a bit more stable than it has been. Second thing is we are ramping our Broadwell-E server in the second half of the year, and so we expect strong demand, and we typically see an ASP uplift as people buy up in the stack with these new server systems coming out. We have customer signals that just indicate that there's a second half seasonal buying pattern kicking in a bit as well. Those three things are the second half keys.
If you take a look at a broader view of this, just overall our view of the data center and really the cloud continues to be that the cloud's going to continue to expand. It's going to be driven by the many machines that connect to the cloud, that drive orders of magnitudes more data than what the average human creates today, and that's getting to the cloud. There's the short versus the longer-term point of view.
Great.
Yeah.
Perfect. For my follow-up, you mentioned you started shipping the 7360. Was that material to Q2, or will it be material to the second half, and any comments on profitability there?
I'm not going to talk specifically about the 7360, because our policy is let our customers announce any design wins that they want to announce when they want to announce them. In terms of the overall impact on financials, when I look at the second half, you'll note that we have an above-seasonal growth rate in there. The biggest driver of that is what Brian just talked about in the data center. That's the biggest driver. We were below the average annual growth rate in the first half. We expect to be above that low double digits in the back half, driven by the three things he talked about. In particular, the cloud buying patterns that we've identified with some of those large customers. Behind that, we see an improvement in revenue in memory and an improvement in revenue in IoT.
Those are the big drivers as we go into the back half.
Thanks, guys.
Thank you. Our next question comes the line of John Pitzer of Credit Suisse. Your line is now open.
Good afternoon, guys. Thanks for letting me ask the question. Stacy, you did a nice job relative to the full year guidance, giving us your expectation for PC demand for the year. I'm just curious, as you look into the third quarter, the guide is sort of the high end of normal seasonal, as you talked about. How would you characterize your view of the PC business going into Q3, and to the extent that it's still a subdued view, is this really confidence in data center, or what other parts of the business would drive above seasonal or high end of seasonal? Sorry.
Hey, John, how about if I start on just the view of the PC, especially into the third quarter, then Stacy can get into where is the above-seasonal numbers coming from and all. I think if you look at the second half, we already said that Q2 ended up being a little bit better than what we had anticipated, and we had built the year at the high single digits. If you take a look at Q2, it ended up coming at the mid-single-digit decline. We've tried to be relatively cautious as we look out into the rest of this year and built the year-end forecast around that high single-digit number set for decline of the PC. We are carrying momentum out of Q2, so there's still data that needs to be collected on how it looks.
Right now, we've maintained our cautious view of the high single digits.
Just to translate that into seasonality against the backdrop of what Brian just said, which is a high single-digit decline in the PC business, we're expecting that segment to play out more or less seasonally, John. We tend to do a little bit better than that because of mix and ASP, but from a unit standpoint, should be fairly seasonal. The driver is what I just talked about. The data center growth rate as we move into the back half should be significantly higher than what we saw in the first half based on stabilization in enterprise, what we see with the cloud customers, and then some ASP uplift as we ramp 14-nanometer Broadwell servers into that product mix.
That's helpful, guys. Maybe for my follow-up, just looking at the memory business, as you guys characterized, tough quarter in the June quarter. If you annualize kind of the operating loss, it's a fairly large number. I'm just kind of curious, have startup costs there peaked? Do we have another couple of quarters of startup costs going up in memory? I guess more importantly, as Dalian ramps and as you think about 3D XPoint, how should we think about kind of your profitability goals longer term and kind of what the fall-through on that business looks like as revenue does start to accelerate?
There's a short-term and a long-term component to this. In the short term, as I think about the second half, in kind of rough math, I'd expect a kind of consistent loss in the second half to what we saw in the first half. I think startup costs will be slightly higher. We'll see the first production costs play through on 3D XPoint, which as you know from watching us over the years, those first production costs tend to be fairly high in any factory when you're starting it up. Offset by the underlying existing NAND business, I think it's a bit better as we go into the back half. That's the short-term answer.
To your longer-term question, I'll turn this over to Brian a bit to talk about the technology, I think the combination of 3D NAND and the cost structure we're going to achieve, then the disruptive nature of 3D XPoint, we should have a very good value proposition and a very good overall profit position for the business.
Yeah, John, I would just echo what Stacy said. We're just now starting to ramp our 3D NAND. As we go through this back half of this year and into next year, it's really starting to ramp up. We think large cost advantages and good performance position there. Then, as we said, 3D XPoint SSDs start to ship at the end of this year, 3D XPoint DIMMs next year. These investments that we're making this year, which we've talked about, are really playing forward those two technologies. I'm still very bullish on the long-term prospects. The units and the gigabytes continue to grow. Our cost structure gets better and better as we go through the back half and into next year. Then 3D XPoint, as we said, will really in our minds change the whole memory storage architecture.
Thank you.
Thank you. Our next question comes from the line of Joseph Moore of Morgan Stanley. Your line is now open.
Great. Thank you so much. I wonder if you could talk about first the data center, the growth. You're mid-single-digit growth for about three consecutive quarters, I know you had higher expectations for that a few quarters ago. Can you talk about, is that entirely an enterprise phenomenon, or is that a timing issue around cloud? Just how should we think about the last few quarters of DCG?
I'd say so overall, relative to the expectations we had at the beginning of the year, let's say, it's primarily enterprise-driven. There's some, what we commonly use is lumpiness quarter on quarter. To how you asked the question, we actually weren't surprised by the Q2 results. In fact, they came in right in line, actually just a hair above what our internal forecast was. We have, I think, pretty good insight into the large cloud customers, we had some good insight into the buying patterns as those customers went from Q1 to Q2. Now as we look at the back half, we see several purchasing cycles kicking in for some of the large guys. We expect that the cloud piece will accelerate as we get into the back half.
Great. That's helpful. Thank you. With regards to the balance sheet, I noticed both day sales outstanding and days of inventory bumped up a little bit, I hadn't expected that. Can you just talk about what drove that?
Those are two different things. I'll just start with the easy one, days sales outstanding. I'm not seeing anything unusual there. It's just the amount of net billings at the end of the quarter. The average day paid and all of that still looks really healthy. I'm not seeing anything unusual there. On the inventory side, as I talked to you last quarter, we ended Q1 higher than where I wanted to be. Our yields got better in Q1. Frankly, they got a lot better in Q2 as well, we undershipped a little bit in Q1 relative to the units that we were expecting. We took some actions in the second quarter to start bringing inventory levels down. What you see inside of inventory, it was kind of flattish, right in line with what we expected Q1 to Q2.
It was a remixing. You see some more expensive server parts and some Skylake parts going up, and then you see some of the older generation CPUs going down. As we get into the back half, we'd expect inventory to click down and be down pretty meaningfully by the time we get to Q4.
Great. Thanks so much.
Thanks, Joe.
Thank you. Our next question comes from the line of Harlan Sur of J.P. Morgan. Your line is now open.
Good afternoon, and thanks for taking my question. PC gaming has been a bright spot for the team with growth in the double-digit year-over-year range for the past number of quarters. Your desktop ASPs are up again in Q2, maybe due to the strength here. Seems like the graphics guys are rolling out some new products. There appears to be a good pipeline of new games for the second half. I guess the question is, did the PC gaming segment continue to drive double-digit growth for the team in Q2? How do you see that trending into the second half?
Sure. If you take a look at it, there was three-ish, maybe four-ish major segments of the PC that did better and continue to do better than the rest of the segment and just kind of overall. Laptops, mobile PCs continue to do better, and they did better in the second quarter. Two-in-one devices, specifically, are doing very well and continue to grow in double-digits. As you said, we often call the enthusiast gaming. You see our K SKUs in there, you saw us also announce the X SKU, which is our new 10-core system that has been selling much better than even we anticipated. Yes, gaming and enthusiast continues to grow at a double-digit rate.
Great. Thanks for the insights there. On the deceleration in the IoT business in June, you talked about an inventory burn. Sorry if I missed this, what vertical was that focused on? Does the team expect re-acceleration on a year-over-year basis as you move into the second half of the year? If so, what verticals are going to be driving the growth?
Yeah. Let me take the revenue question, and I'll let Brian take the what we see in the verticals part. It was not one vertical. If you just look at it, I think our customers got out a little ahead of their skis in the first quarter. If you remember, we had a 21% or 22% year-over-year growth rate in Q1. There was a little more inventory out there than we anticipated, and that took some of the headwind as we started Q2. To your question on the back half, we do expect a re-acceleration. I had said at the investor meeting that we expect double-digit growth in excess of what we had achieved last year. We still expect that. We had a strong Q1, an inventory burn in Q2, we expect a strong Q3 and Q4.
From which verticals? The verticals that have been the strongest growing for us, especially in recent, has been industrial and in the security video-type applications. Those have been the 2 real growth. We have a lot of, I'll call it longer-term growth vectors. Retail is a longer-term growth vector. You saw, as we mentioned in the call, the automotive ADAS section in the announcement with BMW. There's several other programs in that space as well.
Great. Thank you.
Thank you. Our next question comes from the line of Stacy Rasgon of Bernstein Research. Your line is now open.
Hi, guys. Thanks for taking my question. I just wanted to verify on the DCG, I guess, growth target into the back half. It sounds like you need enterprise to keep getting better into the second half, as well as for the, I guess, the cloud, and comm side of DCG to significantly re-accelerate to get there. Cloud was up 9% year-over-year. It's good, but that's a significant deceleration. Can you sort of give us a feeling for what you need for enterprise into the back half and what drove the deceleration in the high-growth parts of this business into Q2?
Yeah. I'd say I characterize enterprise in the back half consistent with what we saw in Q2. If it stabilizes at that rate, that gives us the ability to grow to the levels that we're projecting. For the cloud, yeah, you're right, and it's what I said earlier. We had actually forecasted a pause in purchasing based on what we knew of the customer, the big cloud players' ordering patterns. Based on the signals we're seeing from them, we do expect a re-acceleration in the back half to something more consistent with what we were seeing through the last couple of years.
What gives you confidence that enterprise will continue to stabilize and that what we saw in Q1 is not a one-quarter blip, given it's been down pretty meaningfully for the last few years, pretty consistently?
Well, it's what we saw in Q2.
In Q2, I mean.
Yeah. We saw it down slightly. Again, it's just what we see of the big enterprise customers signal to us in terms of what they want to purchase.
Thanks, Stacy.
Thank you. Our next question comes from the line of Ross Seymore of Deutsche Bank. Your line is now open.
Hi, guys. Thanks for letting me ask a question. Given that dynamic in the cloud and the enterprise side within DCG, do you still expect a crossover in the percentage of revenues that those two generate in the back half of this year? Is it pulled forward a little bit, or pushed out, given the dynamics that you've described so far?
Sorry, crossover from what to what, Ross?
When you said, I believe at your last analyst meeting, you talked about the cloud revenue, the percentage of DCG would be somewhere in the mid-30s, and that would cross over the enterprise-generated revenue in that, and that was therefore viewed as a point of acceleration for the entire group. I'm wondering, given the slower growth year-to-date in that business, if that crossover is going to be achieved in the time you expected.
No, I'll be honest, I haven't looked at the data from that lens. What I said, let me just take you back to the answer that I gave, I think it was to Joe's question earlier, that if you look to the expectations we had at the beginning of the year, the enterprise is a bit weaker than we thought overall for the year. I think cloud will be in line with what we thought. My guess is that's still true. Which quarter it happens may be a jump ball, but my guess is it's still true. I'd like to have to actually go look at the data, and I can give you a crisper answer next time we talk.
Okay. I guess as my follow-up question, sticking with DCG then, is on the gross margin side. It sounds like you're expecting some pretty significant growth, not only in the unit side of the equation in DCG, but also in the ASP side. Then you also talked about some of the ASP benefits that might happen in CCG, at least in the PC portion of it. If I put that all together, I'm a little surprised that either mix or ASPs, neither of those are mentioned in your gross margin reconciliation for it to go to 62% in the third quarter. Can you talk a little bit about how those dynamics fit into your gross margin?
First let me start with your question on data centers. There's a lot in there. I do expect an operating margin improvement in the data center as we work through this year. I think we were around mid-40s in the second quarter, and our historical range has been closer to 50. As I expect it, that ticks up some as we move into the back half. The big driver there is costs. The early production on the Broadwell server, it's fairly expensive. Costs come down as we get into the back half. As you said, we do expect some ASP uplift in the back half. Overall, for the company, what we're seeing, I'll just use Q3 as the anchor point. We're seeing some good news in Q3 associated with higher volumes.
The big offset there is we're seeing 10 nanometer startup costs going up pretty significantly in the back half. That may be the piece that you're missing in the equation in Q3. That continues into Q4, by the way. We see a little bit more startup costs. If you do your algebra, you're probably coming up with a gross margin for Q4 that's about 62%. Think of that as some good things happen, but we have, again, an increase in startup costs, and we have some of the Dalian costs with the first production of 3D XPoint that also kick in a bit more in Q4.
Great. Thank you.
You're welcome.
Thanks, Ross.
Thank you. Our next question comes to the line of Chris Caso of CLSA. Your line is now open.
Yes, thank you. Just a clarification on some of your earlier comments on inventory and PC. You talked about the customers taking down the inventory a little bit less than you had expected in Q2. Can you clarify why they chose to do that? I guess with the inventory levels right now, where do they stand relative to where you'd expect them to be into the third quarter?
Yeah. Overall, we continue to see inventory levels as being very healthy. I think that the worldwide PC supply chain has moved towards a stance of being fairly lean and fairly cautious, and we see that continuing. You would typically see an inventory burn in the second quarter. It was a little bit less of an inventory burn than what we expected. I think it goes back to Brian's comments that he made at the beginning of the call, that, from their perspective, the PC market was a little better in Q2. I think they probably just brought down inventory levels a little bit less.
Okay. As a follow-on to that, maybe you can clarify what you consider to be normal PC seasonality in a third quarter, because I think that's been changing over the past several years. Last year, the build into the third quarter was a little stronger. Obviously, there were some product launches there. How do you characterize the build this year relative to what we've seen last year and the year before?
Yeah. I would say for the company, you would expect to see the seasonality or revenue seasonality for us that's in the high single digits as we go from Q2 to Q3. Our guide is a little higher than that for the reasons that we've been talking about. PC market is probably in line with that, maybe just an ooch higher in terms of the overall PC TAM.
Great. Thank you.
Thank you. Our next question comes to the line of Matt Ramsay of Canaccord Genuity. Your line is now open.
Yes. Thank you very much. Good afternoon. Thanks for taking my question. I guess it's kind of a follow-on to the DCG question that Ross had brought up in terms of margin. The DCG business is obviously going to diversify itself some going forward with some of the new products that you're introducing and the networking business taking off. The operating margin percentage is down fairly sharply year-over-year, I would expect that to ramp back up some as revenue re-accelerates. Maybe, Stacy, you could talk about what the long-term margin structure looks like from an operating margin perspective in DCG as the business diversifies some. Thanks.
Sure. We'll talk about that in more depth at the investor meeting, but I'll give you some off-the-cuff comments here. First, to just put it in perspective, the operating margin percent decline that we saw in Q2 had nothing to do with the new products, or the first order, the new products weren't driving it. It was the 14-nanometer cost of Broadwell server. We get in the back half, we're going to see an ASP impact associated with that because we think that the performance that that product brings enables customers to get a better value proposition by buying a richer mix. That's why I'm pretty confident when we get in the back half, we see the margin snapping back towards that 50% that we've articulated as our long-term goal.
We'll talk more about the mix of products and whatnot at the investor meeting, but the one thing I'd point you to is, remember, something like networking, which is Atom-based server product that's going in, while it does have a lower ASP than the average within DCG, it has an ASP that's actually been going up, and it has a very different cost structure than a Xeon does. I wouldn't just immediately assume that because it's a lower price segment of the market, that it's a margin-hindered segment of the market, because that's not the case.
Great. Thank you. That's really helpful. I guess to follow up on that, sticking on margins, maybe you could talk a little bit about how adding the third chip on 14-nanometer may affect margins going into 2017, in the PC business to offset some of the 10-nanometer ramp costs you had talked about. I think it's a little bit of a different dynamic than we've seen with the business traditionally with the tick-tock approach. Any sort of broader comments you could give there about margins would be really helpful. Appreciate it. Thanks.
Yeah. Let me take the margin side, and then I'll let Brian come in over the top to talk a little bit about the roadmap and how we think about that from the customer's perspective. First off, we haven't forecast 2017 margins, so we'll give you the first glimpse of that when we get to the investor meeting, and then we'll put a formal stake in the ground when we get to next January. I would say that there's nothing that I'm seeing in the overall roadmap that for me is a significant headwind as we go into 2017. We'll give you a lot more insight on that in a few months, but I'm not seeing anything that has me worried as we go into 2017.
Yeah. I guess what I would talk about is Kaby Lake. One of the things we've learned on 14 nanometers is how to make meaningful performance improvements both in the silicon and then with the silicon combined with the architecture. We said we already started shipping Kaby Lake to our customers and OEMs. We're seeing meaningful performance across all of the various SKUs of Kaby Lake relative to Skylake. Kaby Lake is built off a Skylake core. As a result, the die size doesn't significantly grow. There's no driver in the silicon itself to shift the margin structure of this product. We're able to get the performance, and feature enhancements, with relatively small silicon increases, but good improvement on the raw silicon technology itself.
There's not an intrinsic driver that should say, "Die size got twice as big, so margins are cut." There's nothing like that.
It comes in on a process technology that's mature with healthy yields and healthy cost structure. From that perspective, you get a nice performance boost at a good cost structure.
Thank you very much.
Thanks, Matt.
Thank you. Our next question comes from line Romit Shah of Nomura Securities. Your line is now open.
Yes, thanks. Brian, I know that you remain cautious overall in the PC market, when you look at some of the third-party data, it definitely seems like North America was better. Asia, Latin America, a little mix. I'm wondering, has your view, at least on a regional basis, changed at all?
my regional view would say that North America and Western Europe has been stronger for us for some period of time, and it continues to be the stronger segment for us. As you said, South America and Latin America continues to be weak. We see weakness in Asia, but it did get a little bit better than in the past, and that combined with North America were the two drivers that made Q2 perform better than what we'd modeled in our high single digits. As I said, I'm being relatively cautious in this and making sure that we put an estimate out there that we are very comfortable with, and that's why we've stuck with our high single-digit view of the year.
The other thing that's been reported is that commercial models have seen some momentum. Windows 10 has been a catalyst. Your view on commercial enterprise, has that improved at all over the last 90 days?
Yeah. we're hearing that same thing from our customers. As we go out and talk to CIOs, we're hearing the same thing. Those cycles, though, can sometimes take-- You'll hear them, and then they sometimes can take some time to really kick in. again, we kind of built that into the cautiousness of the second half and making sure that we know what we're going into the second half with. we are hearing similar things around the enterprise conversion. it's a comfort with Windows 10, ability to make that transition, and wanting to do it on new hardware.
Great. Thank you.
Thank you. Our next question comes from the line of C.J. Muse of Evercore ISI. Your line is now open.
Yeah, good afternoon. Thank you for taking my question. I guess first question on XPoint. You talked about stronger growth expectations looking out over the next couple of years, and just curious if you could share what conversations you've had with customers, what use cases you've uncovered, and any thoughts in terms of sizing the market.
Yeah. We have pretty widespread guesstimates and models right now on sizing the market because we're still really learning. We've actually started to ship some sample units to customers already, to let them try out and start to learn. Those are to the big cloud service providers, is mostly who we're sending those to. You're going to see it enter as SSDs. You'll see those SSDs, both enterprise class SSDs and also commercial consumer-type SSDs. We've demonstrated in several live demonstrations, anywhere from 5 to 7, 8, 9, 10x improvement in performance depending on the workload, through those SSDs. You're going to see those be the first implications. When I think about where the big volume will come from, I think it will come in that DIMM form factor. You're just going to see it in cloud applications.
Everything from machine learning, big data, any place where you have memory intensive and where you could do in-memory applications, the 3D XPoint is going to be nicely configured for that. It allows you to bring large amounts of storage-like data into a memory-like performance. That's the real key here. I also believe you'll see it in consumer devices. You'll see laptops and devices like that. Gaming machines, we think it'll have gaming applications where you can preload, in a cache-like environment, the next level of your game. It loads almost instantly as you transition within the game. There's going to be a variety of those, and actually, the more we go and start to play with it, start to give it out to customers, the more types of applications and workloads we're finding that it can be used for.
That's very helpful. I guess as a quick follow-up, Stacy, can you provide an update on your targeted capital structure here and at what point with net leverage, net cash, we should start to see more aggressive buybacks?
Sure. As I communicated, I think at the investor meeting or in the last six months, our goal is to get back to net cash zero. At one point, I had predicted that that would happen in the back half of this year. The combination of business levels being lower from what we thought at the beginning of the year, as well as some pretty significant restructuring charges, will push that out. That'll happen sometime in 2017, although I do think we'll make good progress towards net cash zero as we move into the back half. Today, our net debt level is right around $5 billion. We'll take a big chunk out of that as we move into the back half. That said, we're still generating excess free cash flow. You can see that even in the first half, even inclusive of the restructuring charges we did.
You can see we did, I think, $1.6 billion of buybacks in the first half. We have a dividend yield that, as of yesterday, was about 3%. I think you see us executing to the priorities that we had articulated.
Thanks, C.J. Operator, I think we have time for two more questions.
Thank you. Our next question comes the line of Vivek Arya of Bank of America. Your line is now open.
Thanks for taking my question. For my first one, Brian, I'm curious that there are expectations of Intel becoming successful with your 4G modems in the back half. I'm curious, how's the longer-term visibility around sustaining growth in the business, and just the timing as to when you can bring those products from foundry to your own fabs?
Sure. Let's just talk about modems. Really, to stay on a leading-edge modem, you need to have a yearly cadence of modem technology. We have just got a yearly cadence laid out. We had the 7260 last year, we have the 7360 this year. We have a series of modems out. We've got them built out and planned for the next several years. We haven't publicly stated when we'll bring it inside, clearly, we plan to. We'll do that when the right point of intercepting the roadmap, and getting the right performance off the 14 nanometers is required. Right now, I'm more concerned about getting the leading-edge momentum going for us with the 7360 and then the follow-on in 2017, really showing that we are a world-class modem company.
Got it. As my follow-up, staying on the DCG team, are you still comfortable this can be a double-digit growth business over the longer term? I guess as part of that, what role does competition play into it with all the recent noise around SoftBank buying Arm and presumably putting more resources into it? I understand there are no near-term implications, just longer term, how do you think about growth and competition in DCG? Thank you.
Sure. Let me try and talk about why I'm so confident in growth. We can talk about competition, and then we'll see if that does not answer your question. When I think of the cloud, the cloud that we have today is really built on the backs of people. It's your Facebook data, it's your Salesforce data, it's your Twitter data. It's all data that is really across the devices that we pretty much handle day to day. The current estimates are, if you look out into 2020, that average person will generate about a gigabyte and a half a day of data off those devices. Those are going to be all your posts and pictures and all that kind of information. If you take a look at the average autonomous car in 2020, the estimates right now is it'll throw off about 40 gigabytes a minute of data.
If you take a look at the average autonomous drone doing some kind of scan, looking for somebody lost in the forest or scanning a mine, it's going to throw off about 20 gigabytes a minute. If you take something like our Replay Technologies, that is filming in virtual reality a basketball game or a football game, it's throwing off 200 gigabytes a minute right now. As we continue to refine the accuracy of that number will likely just grow. It's that growth in data and the need to both process it at the edge and then through the data center and into the cloud to be able to store it, to be able to apply machine learning to all of those applications. Those all tell me that the cloud's going to continue to grow. It's going to be lumpy.
These guys don't build out their data centers in a linear fashion. They go, they build out a big chunk of over capacity so that they can go then sell that and have expansion space, and they don't build for a while. I know people worry about, well, is it slowing down? It's these trends in data that tell me, no, it's not slowing down over the long term. What you're really going to see is just the buying patterns and the build-outs of the various structures that are going up. As far as competition, there's always going to be competition in this market. I expect it. That's okay. We think of ourself as competition.
In fact, we are built on a model that says we have to build a continuous improvement of our product, such that we're replacing ourself with a better cost per performance model over time. We know that just even if there was no competition, the competition is we've got to build a product that's better and drives replacement as well as growth. I look at the competition as it's welcome. It keeps us better. It's always been out there. There will always be somebody out there. Really what we have to do is build products that are so competitive that people want to replace our products with our new product. That then is one of the best models to use for making sure you stay ahead.
Thank you.
Thanks, Vivek. Operator, can you please go ahead and introduce our first questioner, or last questioner, excuse me.
Thank you. Our last questioner comes from the line of David Wong, Wells Fargo. Your line is now open.
Thanks very much. Just following up from the earlier answer, can you give us some feel for your attitude to total amount of debt? You have a goal of net cash neutrality, but do you have an upper limit on the total amount of debt you're willing to carry? Might you choose at some point to repatriate overseas cash to bring down debt or pay dividends and stock purchases?
There certainly is a limit to the amount of overall debt that I'd feel comfortable with, but it's not something that we've articulated externally, and it'll vary by the size of the company and the cash flow. To my philosophy of just generally, I'm not a believer in taking on debt to do stock buybacks. You really have not seen us do that. Obviously, it's a board decision, but I'm just sharing with you my view of it, of I wouldn't take on debt in order to do buybacks or to do a special dividend or anything like that.
Great. Thanks very much.
You're welcome.
David, did you have a follow-up? Sounds like you're all set, David. All right. Thank you all for joining us today. Operator, please go ahead and wrap up the call.
Ladies and gentlemen, thank you for participating in today's conference. That does conclude today's program. You may all disconnect. Everyone, have a great day.