Good day, ladies and gentlemen, and welcome to the Intel Corporation fourth quarter 2016 earnings conference call. At this time, all participants are on a listen only mode. Later, we will conduct a question and answer session, and instructions will follow at that time. If anyone should require operator assistance during the conference, please press the star, then a zero on your touchtone telephone. As a reminder, this conference call is being recorded. I would now like to introduce your host for today's conference, Mr. Mark Henninger, Head of Investor Relations. Sir, you may begin.
Thank you, Chanel, and welcome everyone to Intel's fourth 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 Bob Swan, 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. A brief reminder that this quarter we have provided both GAAP and non-GAAP financial measures.
Today, we will be speaking to non-GAAP financial measures when describing our consolidated results. The CFO commentary and earnings release available on intc.com include the full GAAP and non-GAAP reconciliation. Finally, I'd like to remind everyone that we'll be hosting our annual investor meeting here at our Santa Clara headquarters on Thursday, February 9th. If you have any questions about the event or logistics, please contact investor relations. With that, let me hand it over to Brian.
Thanks, Mark. I'd like to cover three things with you today before handing off to Bob. A brief review of our results for the year, an update on the transformation of the company, and a look ahead at 2017 and Intel's future. First, the review of our results. Q4 was a strong finish to a record year in which we grew revenue 7% and net income 9%. The Client Computing Group, Data Center Group, Internet of Things, and Intel Security all grew in 2016, with DCG and IoTG setting full-year revenue and volume records. The acquisition of Altera added three percentage points to our overall growth rate. In the data center, cloud service provider revenue grew 24%, while enterprise revenue was down 3% for the full year. Additionally, we are very excited that comm service provider revenue grew 19%.
Across all of our customer categories, adjacency revenue grew an incredible 21%, with strength in Ethernet controllers, Omni-Path Fabric controllers and switches, and network ASIC. IoT revenue was up 15% for the full year, driven by strength in the video, retail, and industrial segments. In Q4, we launched new Apollo Lake and Kaby Lake product, and we won key designs in automotive and video. Our memory business finished the year with record quarterly revenue, while full year revenue was down 1%. This was an investment year for the memory business, but NSG operating margins improved meaningfully in the fourth quarter. We're now shipping 3D NAND from our Fab 68, and we just qualified our first 3D XPoint-based Optane SSDs, which we expect to ship for revenue in the first quarter. Our 3D XPoint memory DIMMs are sampled to data center customers.
The Client Computing business achieved impressive results, driven by strong execution and higher ASPs as customers bought a richer mix of Intel Core product. This was also a breakout year for CCG's wireless communications product line. Our 7360 LTE modem ramped into high volume, and we shipped record Wi-Fi units. I'm very proud of the Client Computing business and what this team has delivered in 2016. Our strategy of delivering consistent product leadership, segmentation, and differentiation paid off. We achieved record i7 units and overall core mix in 2016. Client Computing revenue was up 2% for the year, despite a declining PC market. The Programmable Solutions Group was up about 7% on a non-GAAP basis over Altera's 2015 results. PSG saw strength across many segments, with particular strength in compute and storage.
The Programmable Solutions business sampled the industry's first and only 14 nanometer FPGAs this year, known as Stratix 10. This product line has the largest demand pipeline in Altera's history. The integration of Altera into Intel was an important milestone in what was a transformative year for Intel. This team did a great job of meeting integration objectives while continuing to deliver new products and grow the business. In 2016, we took other important, and in some cases difficult, steps to position the company for future success, improve the alignment of our resources to our strategy, and accelerate our transformation to the company that powers the cloud and billions of smart and connected devices. Our restructuring initiative focused on our investments on the product and technologies that will fuel our growth.
That work, combined with improvements in 14 nanometer cost, drove a 30% operating margin improvement in our client business. Our decision to establish McAfee as a separate independent company was another transformative move. This transaction will give McAfee the flexibility to invest independently, tighten our focus, and allow Intel to share in McAfee's future success as the market demand for world-class security product continues to grow. Finally, looking ahead to 2017 and Intel's future, I'm confident we're making the right investments to compete and win, not only in the segments where Intel is strong today, but also in new areas that are poised for growth, driven by the emerging flood of data. By 2020, the average person will generate about one and a half gigabytes of data per day, while smart and connected devices of the future will produce data at many times that rate.
Autonomous cars, for example, will generate about 4,000 gigabytes of data each day. The resulting explosion of data is creating tremendous opportunity. Data alone isn't valuable. It's the transmission, aggregation, and analysis of the data that results in value and impact. Intel will play a central role in those steps because our products are key to turning raw data into high-value insight and information. Our investments in advanced research and development are making this all possible while significantly expanding our TAM at the same time.
These investments are creating opportunities in segments from autonomous driving, where we are uniquely positioned to be the compute engine in the vehicle and the data center, to 5G, where we are building on our momentum in 4G to establish leadership, to artificial intelligence, where we are providing the industry's most complete range of products to accelerate all AI solutions from the edge to the data center. We're already seeing signs of progress. In autonomous driving, we are winning key designs like BMW, Delphi, and Baidu. In 5G, we are leading in the definition of standards, prototyping, and field trial. Our network business saw strong growth as infrastructure moved to Intel architecture in anticipation of 5G. Our design wins and network virtualization innovation position us for leadership share in the wireless access market.
Other proof points of our progress include important partnerships with AT&T, SK Telecom, Korea Telecom, and Verizon, and the announcement earlier this month of the industry's first global 5G modem. Finally, in artificial intelligence, we believe we have the industry's strongest product portfolio. Intel processors power well over 90% of servers deployed to support machine learning workload, and we are winning the vast majority of AI solutions based on strong product performance and customer value. Wrapping things up, I'm very pleased with Intel's performance in 2016. We have important work to do in 2017 as we continue to transform the company. The progress we've made leaves me increasingly confident in Intel's growth and Intel's future. And with that, I'll hand it over to Bob.
Thanks, Brian. The fourth quarter was an outstanding quarter. Revenue was up 10%, operating income was up 11%, and EPS was up 4% year-over-year. Revenue set an all-time record of $16.4 billion. Operating income was $4.9 billion. Fourth quarter operating margin was 30%, flat year-over-year. Gross margin at 63% was down two points, primarily driven by a couple of one-time events related to product warranty cost and long-term IP agreements. Direct spending came in at $5.4 billion, up 4% year-over-year and down two points as a percent of revenue. Earnings per share of $0.79 was up $0.03 from a year ago. The Client Computing Group had revenue of $9.1 billion, up 4% year-over-year. During the fourth quarter, the worldwide PC supply chain remained healthy, and we saw some inventory burn during the quarter.
Client ASPs were up 7% year-over-year, and core mix was at an all-time high as a result of the success of our segmentation strategies and strength in gaming and high-end systems. This segment had another quarter of significant profit growth, with operating profit growing 30% from a year ago as the business continues to benefit from lower spending, richer product mix, and continued improvements in 14 nanometer unit cost. The Data Center Group had record revenue of $4.7 billion, up 8% year-over-year. In the fourth quarter, we continued to see robust growth in the cloud and comm service provider segments of the business, which both grew approximately 30% year-over-year, partially offset by a 7% decline in the enterprise and government segment over the same horizon. The Data Center Group had operating profit of $1.9 billion, down 14% year-over-year.
Operating margin was impacted by the two one-time items I referred to earlier and the ramp of 14 nanometer on our server products, which we expect to generate continued cost improvements over time. Our Internet of Things business achieved record revenue of $726 million, growing 16% year-over-year, driven by strength in both retail and industrial segments. Operating profit for the business was $182 million, up 37% year-over-year. Our memory business had record revenue of $816 million, up 25% year-over-year, with strong demand for data center SSD solutions with demand signals outpacing supply. We've made great progress ramping Fab 68 with yields and unit costs well ahead of expectation. This segment had an operating loss of $91 million, largely driven by costs associated with 3D XPoint and start-up costs for our memory capacity.
The Programmable Solutions Group had revenue of $420 million and operating profit was $80 million. Our Intel Security Group business had revenue of $550 million and operating profit of $103 million. Turning to the full year 2016, revenue grew 7%, operating margin grew 11%, and EPS grew 9%. Operating margin percent improved by one point while gross margin was approximately 63% flat to 2015. Spending was $21 billion, or 35.4% of revenue, down one point from 2015, primarily from the restructuring programs we began earlier in the year. Operating profit for the year was $16.5 billion. Earnings per share for the year was $2.72, up $0.23 from the prior year. In 2016, the business generated record cash from operations of $21.8 billion. We purchased $9.6 billion in capital assets, paid $4.9 billion in dividends, and repurchased about $2.6 billion of stock.
Total cash balance was $17.1 billion, down $8.2 billion. Total debt was $25.3 billion. Our net cash balance, total cash less debt, and inclusive of our other longer-term investments, is approximately negative $2.3 billion. Now let me turn to guidance. First, some context. First, our guidance assumes a stable macroeconomic environment, but we have taken a more cautious view of PC consumption versus third parties, particularly in our outlook for the emerging markets, including Russia, China, and Latin America. Second, for the data center, we continue to expect similar growth rates in the cloud and comms segment, but we are not expecting an improvement in enterprise. This gets us to an expectation of high single-digit growth in the data center business. Third, as a reminder, we have one less week as a result of the inclusion of an extra workweek in 2016.
Last, we have assumed the Intel Security transaction will close in Q2, and our full-year guidance reflects one full quarter of the group's consolidated results. We have lots of work to do to close this out, and in the event the close happens at the end of the second quarter, we would update our full-year guidance by approximately an additional $500 million of revenue and $100 million of operating income. As we look forward to the first quarter of 2017, we are forecasting a midpoint of the revenue range at $14.8 billion, up 7% year-over-year and down from the fourth quarter. This is at the lower end of our seasonal range and reflects an expectation of lower core brand mix and ASP coming off a strong holiday selling period for gaming and other premium PC systems.
For the first quarter, we expect operating margin percent to increase four points year-over-year, gross margins to be flat at approximately 63%, and spending to be $5.3 billion, down 1%. We expect EPS to be approximately $0.65, up 20% year-over-year. Turning to the full-year 2017, we are expecting revenue to be roughly flat. Revenue is expected to grow in the low single digits after excluding the Intel Security Group from both years. We expect operating margin percent to be up one point year-over-year with flat gross margins and direct spending as a percent of revenue down one point versus 2016. Let me provide a bit more detail on our year-over-year direct spending. Our restructuring plans are on track, including reducing our headcount by approximately 15,000 heads and generating gross savings of $1.6 billion.
We are reallocating investments from CCG to higher-growth segments and are continuing to invest in areas that extend our leadership position in Moore's Law and expand TAM opportunities such as memory and autonomous driving. We anticipate the net benefit of these actions to result in an additional point of improvement in direct spending as a percent of revenue in 2017. We expect EPS of approximately $2.80, up 3% year-over-year. In 2016, we had a fairly significant gain from our ICAP portfolio. Our 2017 guidance assumes that we will have gains roughly in line with 2016 levels. The capital spending forecast for 2017 is $12 billion, up $2.5 billion from 2016 as we continue to ramp our memory capacity. In closing, 2016 started out slow but finished strong.
Our CCG business is focused and executing extremely well in a declining market and provides scale that's required to advance Moore's Law, generate significant cash flows, and enables investments for growth. Our growth-oriented businesses were up 15% collectively as we continue to transform the company from PC-centric company to a company of smart and connected devices that power the cloud. We are excited about our plans for 2017 as we continue with our transformation, and we look forward to sharing more details with you at our investor day in February. With that, let me turn it over back to Mark.
Okay. Thank you, Brian and Bob. Moving on now to the Q&A. As is our normal practice, we would ask each participant to ask one question and just one follow-up if you have one. Janelle, please go ahead and introduce our first question.
Our first question comes from the line of Chris Danely of Citigroup. Your line is now open.
Hey, thanks, guys. Just digging in on the segments. The CCG operating margins were, I think, the highest in the last few years. DCG was kind of the lowest in the last few years. Can you just talk about the trend in operating margins for those two segments for this year, given the guidance?
Yeah, Chris, the CCG performance, couple dynamics that are driving the record margins. One, higher ASPs on the strong mix that Brian referred to. Secondly, real strong unit cost. It's been on 14 nanometer for a while, we expect that to continue. Constrained investments in the CCG business as we focus more on invest in heavier in the higher growth businesses. Real good performance, and those are the dynamics that we're expecting in 2017. On the DCG front, good in Q4 and then project forward. Nice ASP performance in the quarter as we migrate from Haswell to Broadwell, and as we think about Skylake in 2017, we expect that to continue. Unit cost in DCG is a little bit higher because we're coming off more great unit cost performance on 22 nanometer.
As we migrate to 14 nanometer, in the earlier stages of that cycle, unit costs are a little bit higher. Then we're increasing our investments in DCG on a year-over-year basis. Dynamics are relatively good. The one, I referred to it a little bit earlier, but in terms of Q4, we had a couple one-time items that really weighed on the profitability or operating margins of DCG that we think are well-bounded and do not expect to continue in 2017. Just quickly, one, we entered into a long-term cross-license agreement and patent purchase agreement, particularly in the comm space in the quarter, and DCG was impacted by a portion of that cost in the quarter. Secondly, and a little more significant, we were observing a product quality issue in the fourth quarter with slightly higher expected failure rates under certain use and time constraints.
We established a reserve to deal with that. We think we have it relatively well-bounded with a minor design fix that we're working with our clients to resolve. Those two one-timers in the fourth quarter weighed on DCG margins. We do not expect that to continue in 2017.
Great. Thanks.
Thanks, Chris.
Do I get a follow-up?
Yeah, please go ahead, Chris.
Okay. Maybe just expand on the PC market. What do you see driving the strength here? How sustainable is it? Any thoughts on PC unit growth for this year?
This is Brian. I can start, and then Bob can add in. If you take a look, Bob mentioned in his portion of the talk of, hey, we've taken a slightly more conservative view of 2017 than third parties for what we see the overall PC unit market as. We had extremely strong, record Core i7, just record core mix in the fourth quarter that as we look at Q1 and we look at 2017, we've factored a little bit of caution into that as well. Those two things put us in the PC market at a unit level in the mid-single-digit decline. That's better than if you went back a year or so ago, we were in the high single digits, depending on how you looked at it and where you counted some of the two-in-one devices.
It is starting to get better, but I don't think we're back into at a zero unit or a positive unit. As Bob said, what we've really been focusing on in that space is how do you make money, how do you sell up, how do you do better business performance in that kind of market? We're comfortable that we can continue that into 2017.
Chris, the only thing I'd add is, we think, as Brian said, our outlook is a little more conservative than the third parties. Our view is that's probably the right posture and the right caution to have as we go into the year. Obviously, the team has done a great job in adjusting its cost structure for a more cautious outlook, then we'll see how it plays out during the course of the year.
Great. Thanks, guys. Nice quarter.
Thank you.
Thanks, Chris.
Thank you. Our next question comes from the line of Joe Moore of Morgan Stanley. Your line is now open.
Great. Thank you. I wonder if I could ask about the DCG commentary of being up high single digits. Obviously, that's a little bit more conservative than the longer-term numbers you had talked about. I guess, as you think about that, is there an element of the cloud spending that you think is causing the enterprise to be weaker? Maybe just give us the quality of underpinnings for that change.
Sure. Let me start, then again, Bob can jump in on this one. As we said, we took a look at the 2017 view on enterprise to be relatively equivalent to what we saw in 2016, which says that enterprise continues to decline. I think that certainly some of that is that it's moving to the public cloud. It's moving to those areas at a faster rate than I think we expected. It's also been a little bit slow about developing private clouds, and we're working with several partners like Microsoft Azure and others around the private cloud segment as well for the enterprise. If you take a look at the long term, we still see this as the growth engine and still getting into that double-digit regime. Remember, for us, enterprise is now less than 50% of our overall data center business.
The areas that are growing even faster or as fast as cloud in most cases are the networking and storage space, which we have very low market share in still, and it's a great opportunity for us. The emerging areas like silicon photonics, Omni-Path Fabric, Rack Scale Design, and 3D XPoint. Those areas are really what we've always forecasted to be the growth engine of the Data Center Group as we go towards the back half of this decade. For us, our view is anytime you're going through a market transition, you're not going to get the cloud to enterprise mix perfect. This is an anomaly right now that we've forecasted, we think, accurately and adjusted for it the way it is. Our long-term growth was actually based on other factors, and we're still very confident in those growth areas.
I don't know if Bob wants to add.
No, I think that's perfect. Nothing to add. Thanks.
Great. Thank you for that. That's great. Separately on the PC market, the ASP growth that you guys saw over the course of 2016, is that strictly mix shift and kind of strength in the higher-end segments as you've been talking about? Any thoughts on your ability to continue that price momentum over the course of this year?
Sure. Yeah, it's mostly, if not in almost every case, all mix shift, and it's our customers buying up. A great example was the K SKUs, the enthusiast, the 10-core systems that we've put out there, and they've far and away exceeded our original sales forecast for people who are out there buying 10-core gaming systems. We do believe that that market, that enthusiast market will continue. We factored a little bit more caution into this as we go into 2017 in the first quarter. Some of that's seasonality. Holiday, people tend to buy a lot of gaming systems. Some of it's just how much more can people buy up, and so how much more growth in ASP can we see? We don't see a decline or anything of the average ASPs.
The only thing I would add is as we think about the full year, the second half comps get a little bit tougher on ASPs because of the strong ASP performance throughout the course of 2016. That's probably the only other dynamic that I would add.
Great. Thank you very much.
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. I guess the first one on OpEx, it's more of a conceptual one for you, Brian. On flat revenues, it's great to see some OpEx leverage on there, I think some people are hoping for a little bit more as you had the restructuring only halfway done and then the McAfee sale that's pending. Can you talk conceptually how you balance the desire to reinvest versus the desire to get the profitability up to your long-term targets?
Sure. I'll let Bob talk about the McAfee financials and how that affects OpEx, because it's a little complex with end of Q1 target there. Let's just talk about it in general. What we always said is that we were going to go through the program that we went through at the middle of last year, which we called ACT, which has continued through the end of last year, 2016, and we said would be completed about the middle of 2017. We always said there'd be some mix between taking that to the bottom line and reinvesting it in those growth areas. As we've done acquisitions like Nervana for artificial intelligence, we want to invest in those businesses now to bring them onto our silicon, to integrate them into our software stack. We're going to make those kinds of investments in key areas.
We said the key areas around data center in general. We have Rack Scale Design, we have 3D XPoint, we have artificial intelligence. Around IoT, we're making big bets around autonomous driving. You see us making the investment in here, the investments with BMW. When you take a look at autonomous driving and why are we doing that, it's around data and data centers again. Remember, every one of those high-definition maps is going to require data centers. They're going to require small data centers at the edge. It's all around, are we understanding and managing how data is going to flow in that system? Memory itself, we're going to go make those investments around 3D XPoint and really bringing that and 3D NAND to market in a big way.
We'll balance between those, and I think you'll see some mix of bringing it to the bottom line. If I can invest and think I can turn that into additional profitability in the future, I'm going to go do that.
Yeah. The only thing I'd add is just to maybe make it a little simple for people is we look at our full-year guidance that we gave, then we extract out three quarters of McAfee. At the macro level, we're looking at low single-digit growth and mid-single-digit EPS growth. That's the year-on-year apples-to-apples dynamic. On direct spending itself, our guidance implies we'll be down roughly half a billion dollars year-on-year, and that's just a function of three things. One, continued benefits from the restructuring actions that we took in 2016 and continue to execute on in the first half of 2017. Secondly, obviously, the direct spend of McAfee goes away. Third is, as Brian mentioned, during this transformation, we continue to make the investments in the higher-growth businesses.
We'll continue to invest in 5G and ADAS, then we'll continue to invest in Moore's Law as we bring 10 nanometer to life in 2017 and continue to invest in seven nanometers. Net-net, the implications of all that for direct spending was one point down as a percentage of revenue in 2016 and another point down as a percentage of revenue in 2017. The only other thing I would add is in terms of the milestones that we employ during the course of the year for these big bets. We'll continue to build milestones in to make sure those bets that we're making are turning out in the medium and long term the way we expect.
That's very helpful. I guess as my follow-up, you talked about the ASPs in answering a prior question. I wondered about the competitive intensity in the PC market. You're taking a more conservative tack than the third-party vendors are forecasting, but your primary x86 competitor is coming out with a new architecture for the first time in many years. I wonder whether it's on the ASP or the unit or the market share side, how you're factoring that into your forecast for the year.
Sure. I would tell you that we always look at this environment and say, there's going to be a competitive risk in the environment, we're always focused on really our own product roadmap and making sure that we have the highest performance product. When we look at 2017, we still believe that our product roadmap is truly the best it's ever been. As we look at the Kaby Lake and as it really ramps up through 2017, where it came out really just at the end of 2016, now we'll ramp with many more SKUs and higher performance products as we go into 2017. We showed at CES the first working 10-nanometer Cannon Lake product, which we're still planning to ship by the end of this year and really ramp into 2018.
We still believe that our roadmap and our leadership will continue to give us the performance that our customers want and desire. That didn't necessarily factor into that more cautious forecast. That forecast was really much more a function of where we think the PC market really is overall.
Great. Thank you, guys.
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 questions. I had a question first on the guidance for next year. Unless I'm doing the math wrong, to get to $2.80, I need a fairly sizable reduction in share count. Is that true? Could you tell us how you're thinking about shares for next year? Are you intending to use the cash from McAfee to buy back shares to get to that number?
In essence, the guidance year-on-year doesn't really anticipate any dramatic change in our share count. I think philosophically, our approach is to offset dilution from our comp-based programs. All else equal, share count relatively flat year-on-year. I think the one thing worth noting is in our ICAP portfolio in 2016, we had a fairly significant gain. What I indicated in the prepared remarks is we expect roughly, in 2017, that gain to be in line with what 2016 generated. I think just in terms of implied share count in our guidance, it's essentially flat year-on-year.
Okay. For my follow-up, I wanted to again dig into the OpEx. You talked about a mix of the restructuring, the cuts versus the reinvestment. If I throw the McAfee cost back in, you've actually got OpEx going up fairly sizably year-over-year, and your employee count is actually up year-over-year, even though you supposedly had a pretty big layoff. I guess, can you talk a little more specifically about exactly what the additional spending is going on? Is it people versus technology versus something else? Where is it going? Do you view those investments as looking to open up new markets versus being defensive in nature, or maybe a mix of both?
Yeah. In terms of the type of cost, maybe twofold. Yes, people, and yes, technology/Moore's Law at the macro level, and that expresses itself in higher depreciation year-on-year. In our guide, our depreciation is up quite a bit, and a portion of that will flow through direct spending. Again, on a more macro basis, at the risk of maybe repeating myself, we are investing more in DCG, and in particular, bringing some of these adjacent products that Brian referred to market. We continue to invest in memory, and particularly the 3D XPoint product, and we continue to invest in IoT. Those three businesses are getting a disproportionate share of the investment because those are the businesses that we've seen really strong growth in 2016, and we're counting on continued growth in 2017.
The second area, again, we talked about this a little bit, we see real opportunities in autonomous driving that play to our strengths and our capabilities. We are making a step function increase in our investments to position ourselves very well for that industry and that market as it evolves. Again, third, 7 nanometer technology investment and the spend associated with building a new pilot line in 2017 is also an additional investment. We're executing on our restructuring programs. We made the tough decisions in 2016 that resulted in roughly 11,000 fewer people as a result of our restructuring program. We're not quite done, I'd say we're on track. At the same time, we're making investments both in technology and people to strengthen and enhance our competitive position in the areas that we see as real opportunities for us.
Okay.
The only thing I would add, Stacy, as more of a blunt answer to your question is, if you take a look at the areas that we're talking about, in almost every case, these are new and expanding TAMs for us. Even when you look at the places where we're going in the data center, and this was my point earlier about enterprise versus cloud. Enterprise is now less than 50% of our total makeup for the Data Center Group. The cloud's growing great, and that will continue. If you look at the areas, the majority of the rest of the growth for the rest of this decade in the data center alone, it's networking and storage where we have very low market share today. We're bringing things like software-defined networking and NFV to those.
That's a growing and expanding TAM as those markets move to Intel architecture. It's going into Rack Scale Design, Omni-Path Fabric, silicon photonics. Those are all, again, new, either nascent or expanding TAM for us. Autonomous cars in the IoT space are new and expanding TAMs for us. 5G is a TAM that's brand new that'll be really being built out over the rest of this decade. Memory, if you look at the large part of the investment we're making at 3D XPoint, which will really re-architect memory and storage and will create a new market in our mind. We believe we're unique in having that technology. To me, when I look at the investment, they're all focused around data. They're all in support of how the data center ecosystem works, and they're all in either expanding or new TAMs for Intel.
That's why I see the growth in those areas. It's not just enhancing the technologies that we already have. We'll do that, but the new investments are really focused on the new areas.
Got it. Thank you, guys.
Thanks, Stacy.
Thank you. Our next question comes from the line of John Pitzer of Credit Suisse. Your line is now open.
Yeah, good afternoon, guys. Thanks for letting me ask a question. Brian, my first question, I want to go back to the DCG ASPs in the December quarter. They were up 4% year-over-year, which kind of reverses, I think, a four or five quarter trend of ASPs going down. I know that you've had some mixed headwinds that have been driving blended ASPs down. I'm kind of curious, what happened in the fourth quarter to drive ASPs up, and do you think it's sustainable? Is this just what we would expect to see the first quarter of the Broadwell launch and then it normalizes going forward? Or how do you think about ASPs from here?
I'll let Bob start with this, and then I'll come in and talk a little bit about the macro view of this.
Yeah. Two dynamics in the fourth quarter where we had higher ASPs. One, the continued transition from Haswell to Broadwell is helpful. As we project forward, the next transition to Skylake we believe will be helpful as well. Those dynamics where we're delivering better performance for our customers, we are able to capture some of that in ASPs, and we saw a little bit of that in the fourth quarter. Secondly, in the comms and network space, which is a share gain opportunity for us in DCG, getting those clients to move up the stack in terms of the high-performance server CPUs is the second dynamic. Both of which we think are helpful as we exit 2016 and go into 2017.
Yeah. The only thing I'd add, John, is as we go back into the second half of this year, do this Skylake transition, and that is a technology that will increase performance and the performance per cost to our customers. One of the largest improvements in a long time, if not ever, on the data center. We expect typically when that happens, people see the value in that and they tend to buy up, they tend to buy the better SKUs. To my view, this trend of higher mix should continue.
That's helpful. Then maybe as my follow-up, Brian, relative to the full-year guide, you are kind of expecting a pretty significant drop in free cash flow this year with the increase in CapEx, and you highlighted that CapEx is going to the non-volatile memory group. I am just kind of curious, given that that business, even though it made some improvements in losses in the calendar fourth quarter, is still in sort of a loss position. I am assuming the higher CapEx is going to be a headwind to getting to profitability. I guess, how do you think about the path to profitability, the longer-term business model in memory, and what it might do to the DCG growth rate longer term if Crosspoint is successful?
Sure. Let's talk about memory in kind of a big picture, John. Then I will let Bob talk a little bit about how the CapEx plays and what our kind of view on CapEx is in this space. But we are in this space for one reason. Because I understand it is a cyclical business that tends to be fairly difficult from a price capacity standpoint. But we believe we are coming at it with two very unique technology. Our 3D NAND technology has some of the best performance and best cost in the market. Our current version of 3D NAND has a 15% price or cost value over the competition. Our next version, second generation, has even higher when you look at it on a density basis.
We believe we're going to be able to bring differential cost and performance in 3D NAND that will give us a unique position, and that combined with our knowledge of the data center should allow us to really provide compelling product for data center SSDs. 3D XPoint is very different in that it's a unique technology that bridges between memory and storage. We believe it can re-architect how big data applications, artificial intelligence applications, where you want large amounts of data being brought up as close to the compute as you can, will really transform not only the architecture of those systems, but the performance of those systems. We've demonstrated on stage, even on client systems, using these types of SSDs on an equal price, you can get five to 7x performance improvement using 3D XPoint as a large memory storage combination.
We're investing purely because we believe we have this differential technology. That's why we're in this business. I think if I didn't have that differential, not sure it's a business that Intel would necessarily be in, right? With that leadership, and that we believe we can sustain that leadership, we believe it's a good business and a good investment. I'll let Bob talk about how long and how we view the capital.
Just to add, John, the CapEx dynamic first, kind of at the macro level, up $2.5 billion year-on-year, driven by two things. One, memory, obviously, but also bringing 10 nanometer capacity online. If I just go down a level to memory, roughly $1.6 billion CapEx in 2016. Expectations, it'll be roughly $2.5 billion in 2017 as we bring the incremental capacity online. As we look at memory specifically in 2018, we think it begins to drop off a little bit as we focus that capacity on 3D NAND and increasingly 3D XPoint.
Perfect. Thanks, guys. Appreciate it.
Thank you.
Thank you. Our next question comes from the line of Vivek Arya of Bank of America Merrill Lynch. Your line is now open.
Thanks for taking my question. Brian, I'm curious, with this new U.S. administration, there's a lot of interest in using U.S.-made products. Since you have fairly sizable fabs here in the U.S., I'm wondering how you can take advantage of this environment, or if you're interested in making a bigger push in your foundry business.
Well, we're always open for business in foundry, and we're always interested. Remember we said our foundry strategy was really to be on the leading edge. We can get paid for our technology, and it really allows us to use our unique differentiation in that space. Beyond that, I would just tell you we've always been proud. It's not a new transition or a new strategy for Intel. We've always been roughly between a little bit more than half to two-thirds of our capacity in the U.S. We're the second-largest exporter in the U.S., and we're proud of that position. Other than that, there's no real shift in our strategy right now.
Got it. As my follow-up, back to DCG, there seem to be two moving pieces. You have this declining but very profitable enterprise part, but a faster growth, but perhaps less profitable networking and cloud and other areas. Is that a fair characterization, and do you envision a point at which the non-enterprise parts become dominant enough so you can actually see an acceleration in DCG back to your traditional double-digit type growth rate?
Let me start, and then we can see. I think, our view is that enterprise will continue to decline. A lot of that is those workloads moving to the cloud. It will get to a point, though, where it starts to stabilize, because there are still workloads that will want to be in a private cloud. At the same time, we believe as the world becomes connected, cloud will grow at a much, much faster rate. I made a point in the prepared remarks where, if you look at the cloud of today, being mostly based on people, the average person will generate about a gig and a half of data a day. An autonomous car, when those things start hitting the road, and we've started to build these data centers for some of the trials we're working with.
You're talking about petabytes of data that you're having to deal with, and 4,000 gigabytes a day off the average autonomous car. You put a couple of those on the road, and you need petabytes of storage to handle that. We do believe that the cloud will move at a faster rate as these connected devices become basically more available. That said, the cloud is becoming bigger than the enterprise. We said enterprise is now less than 50%, and we believe the other areas that will grow, networking and storage, the adjacencies like Omni-Path Fabric, Silicon Photonics, Rack Scale Design, which we're working with our partners on, it really lowers the cost of the system and re-architects how the rack is laid out. 3D XPoint , that will drive the growth for us as we go through the rest of this decade.
We believe when you add those up together, this thing will go back to double-digits. When exactly that is, because we're trying to grow these new nascent areas and manage the decline of enterprise. It's going to be hard to call exactly when, but we do still have a strong belief, and we believe the products are very compelling, that these will drive us to double-digit growth long term.
The only thing I'd add, Brian, is on like-for-like product, ASPs have a tendency to be lower to the cloud service providers. At the same time, the cloud service providers really value performance. In terms of the mix of their product, they'll value performance in the higher-end products more than maybe enterprise as a whole.
Right. Thank you.
Thank you. Our next question comes to the line of Romit Shah of Nomura. Your line is now open.
Yes. Thank you. I just had one question. I noticed that you didn't raise the dividend in January. Bob, I'm curious if your view on capital returns, buybacks, and dividends is different than what Intel's done historically.
Yeah. I think historically, the philosophy around first and foremost, investing in organically and our capabilities, has always been the first priority. That'll be the same. Secondly, we'll continue to look at M&A that'll strengthen our capabilities, so that's no different than the past. Third, in terms of capital returns, our expectation has been, and I think we'll continue to think this way going forward as it relates to dividend, grow it in line with non-GAAP earnings, and have it be roughly 40% of the free cash flows of the company. Those change around the margins over time, depending on the CapEx intensity of the business. I don't expect that to change, and we'll continue to look at how we move the dividend in line with that philosophy.
As I said earlier, in terms of the more holistic capital returns bucket, we will continue to offset dilution, which I think is pretty consistent with what we've done in the past. Third, we have a great balance sheet, and I do think that opportunistically, when it makes sense. Reducing our outstanding float is an opportunity we have as we get closer to the net cash zero position that we've been tracking towards over the last 12 months since the Altera acquisition. Philosophically, no dramatic change, dividend in line with non-GAAP earnings growth, but trying to stay in that roughly 40% free cash flow world. Maybe if opportunities present itself, be opportunistic in share count without limiting our financial flexibility relative to the things that matter most, which is strengthening our business.
Great. Thank you. Nice quarter.
Thank you.
Operator, I think we have time for two more questions.
Okay. Our next question comes from the line of Harlan Sur of JP Morgan. Your line is now open.
Yeah. Hi, good afternoon. This is Bill Peterson calling for Harlan. Congratulations on the nice quarter. Coming back to the storage market, try to understand how you view growth this year in light of obviously good sequential growth in the prior quarter, but also overlaying that with the Optane qualifications that are going on in progress. How should we view that in terms of the incremental growth driver in 2017?
Sure. If you take a look at it, as I said, this is a cyclical market. If you take a look at 2016, started out with an oversupply, came into the back half of the year with an undersupply, really, of capacity. We're entering 2017 with a continued tightness in supply. That makes the pricing stable to better. We expect that at least right now, the estimates are through the first half. It's pretty hard to project out through the second half, we've kind of kept the second half relatively calm and cautious. If you take a look at 3D XPoint, as we said, we've qualified, we've started to ship DIMM samples to the big cloud service guys. Those are targeted for 2018 revenue shipment. On an SSD basis, we'll start shipping for revenue this quarter.
If you take a look at the full year, I think the estimate is it's around 10% of our total revenue is 3D XPoint. It could take off, and it could be a little bit more than that. It could take a little while to qualify some things if it's a little off, but you should think it's around 10% of the revenue, and really ramping much more into 2018. What we're proud of is, you get past that first hurdle of getting the first one to production ready and starting to ship samples to the cloud guys and actually getting ready to start shipping SSDs for revenue to the client devices and all that. We're pretty excited about just getting to that point right now with 3D XPoint.
Just to-
Appreciate the color.
Be more tactical, the full year for memory was down 1%, 2016 over 2015. The momentum that you saw is that supply chain dynamic that Brian highlighted changed. As we began to scale our own capabilities, you saw the strong exit growth of 25% in the fourth quarter. With that, going into 2017, we feel pretty good about that growth rate in the fourth quarter as we enter the new year.
Great. Appreciate the color on that. A question on the programmable solutions. The group has shown sequential declines in the prior two quarters versus your main competitor that's shown sequential growth, albeit modestly. You've discussed taking share, now you have the 14 nanometer-based product. I wonder if you could provide some color on maybe on why that group has lagged in the prior few quarters, more importantly, when we should see the inflection in the business and how to think about growth in that business this year.
Let's talk about that. If I look at 2016, as we showed, we had about 7% growth over what Altera had in 2015. If you take a look at it, there's a couple of big segments that are driving that: telco, data center, in the networking space as well. Those are kind of the big three segments. We started to see as we went through, especially the back half of the year, good connection between our ability to go in we're better together with Xeon and the FPGA as we go into that networking space. As our footprint grows in networking and storage, I think it also gives us an ability to continue to bring both products to those markets. We think in 2016, we actually gained share relative to the competition.
As we look out into 2017, we forecasted again above market growth from what we understand are the current estimates, we believe we'll gain a little bit of share again in 2017. Remember, these markets are slow moving. You don't grow 10% share instantly because the design cycles and design conversions are relatively difficult. It's again driven by data center, networking, the telco industry. With Stratix 10, as we said, largest design enablement in the history of Altera. We're really excited. We believe it brings a performance and a cost to our customers that is truly industry-leading and shifting. We are very comfortable or confident in that. Again, that will really start to ship in the second half of 2017. I think that will be really driving growth in 2018.
Really, if you think about these design cycles, that product will really continue to drive growth for probably the next three years plus, just because these cycles are fairly long.
Terrific, Tyler. Thanks. Congratulations on the quarter again.
Thank you. Chanel, please go ahead and introduce our last questioner.
Okay, no problem. Our next question comes from the line of Blayne Curtis of Barclays. Your line is now open.
Hey, guys. Thanks for squeezing me in. I just wanted to ask Bob on the gross margin, your full year is equal to the Q1. As you look in the back half, I just wanted to make sure I heard it straight. It sounded like 14 on the server side may have some initial yields, and that would be a headwind, but I'm just curious on the PC client as well. Maybe you can just talk about the puts and takes as you ramp in 10, but you still ship a lot of 14. As you look into second half, what are the headwinds and tailwinds to gross margin?
Yeah, the two headwinds, sorry, two tailwinds on the year-on-year are flat. ASP is being a little bit better, and we anticipate that more on the DCG side. Secondly, unit cost being a little bit better, and we anticipate that more on the CCG side. Again, as I mentioned earlier, for DCG, as we transition more from 22 to 14 nanometer, all else equal, that will be a little bit of a headwind in the early stages of yield for server on 14 nanometer. Those two dynamics themselves, good growth, increasing profitability, good earnings. However, they're at a lower margin, so the mix dynamic of those are real on a year-over-year basis. Good ASP, good unit cost, but mix is a challenge.
Year-on-year factory ramp, both 10 nanometer and memory are a headwind for the full year, but it gets a little bit better in the second half. The dynamic to first half, second half, it gets a little bit better in the second half.
Thanks. Just maybe a clarification, the $3 billion that you're getting from the sale, should I assume that that goes to debt retirement, like you have been using your cash flow? I guess you should get, at some point this year, potentially at the end of the year, to that net cash zero. Should we think about buybacks at that point?
Yeah, I think the dynamics of the $3 billion, the intention is that roughly $1 billion comes in line at time of transaction, and that we will provide seller financing in the early stages. We'll only get roughly $1 billion upfront. In terms of then just the net debt position during the course of the year, you can assume that we have some maturities in 2015 that we'll take out, sorry, 2017, that we'll take out. We believe by the end of the year, given those dynamics, we'll be closer to a net cash zero position.
In terms of that gives us the much stronger balance sheet and how we think about that, we'll continue, as I said earlier, invest in business, return capital to shareholders, and opportunistically, whether it's reducing outstanding float, that's one that we'll continue to look at and be opportunistic as opportunities present themselves.
Thanks.
Thanks, Blayne. All right. Thank you all for joining us today. Chanel, please go ahead and wrap up the call.
Ladies and gentlemen, thank you for participating in today's conference. This concludes today's program. You may all disconnect. Everyone have a great day.