Good day, ladies and gentlemen, welcome to the AMD Fourth Quarter 2012 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 assistance during the conference, please press star then zero, and an operator will help you. As a reminder, this conference call is being recorded. I would now like to introduce your host for today's conference, Ruth Cotter, Vice President of Investor Relations. Please go ahead.
Thank you, welcome to AMD's fourth quarter and year-end earnings conference call. By now, you should have had the opportunity to review a copy of our earnings release and the CFO commentary. If you have not reviewed those documents, they can be found on AMD's website at quarterlyearnings.amd.com. Participants on today's call are Rory Read, our President and Chief Executive Officer, and Devinder Kumar, our Senior Vice President and Chief Financial Officer. Lisa Su, our Senior Vice President and General Manager, Global Business Unit, will be present for the QA portion of the call. This is a live call and will be replayed via webcast on amd.com. I'd like to take this opportunity to highlight a few dates for you. John Byrne, Senior Vice President and Chief Sales Officer, will present at the Goldman Sachs Technology and Internet Conference on February 12th in San Francisco.
Rory Read will present at the Morgan Stanley Technology Media and Telecom Conference on February 26th in San Francisco. Our first quarter quiet time will begin at the close of business on Friday, March 15th. Lastly, we intend to announce our first quarter earnings on April 18th. Dial-in information for the call will be provided in mid-March. Please note, non-GAAP financial measures referenced during this call are reconciled to their most directly comparable GAAP financial measures in the press release and CFO commentary posted on our website. Before we begin, let me remind everyone that today's discussion contains forward-looking statements based on the environment as we currently see it. Those statements are based on current beliefs, assumptions, and expectations, speak only as of the current date, and as such, involve risks and uncertainties that could cause actual results to differ materially from our current expectations.
Please refer to the cautionary statement in our press release for more information. You will also find detailed discussions about our risk factors in our filings with the SEC, and in particular, AMD's quarterly report on Form 10-Q for the quarter ended September 29th, 2012. With that, I'd like to hand the call over to Rory Read, AMD's President and CEO. Rory?
Thank you, Ruth. We made progress in the fourth quarter delivering on our commitments, managing expense and cash, and beginning to transform AMD for long-term growth and profitability. Full year results fell short of our expectations as the challenging macro environment resulted in a weaker than expected PC market. As I mentioned on our last earnings call, we are executing a turnaround that will take several quarters. We expect continued choppiness in the PC market in the first half of 2013, we will closely manage the business as we reset, restructure, and ultimately transform AMD. We have the right strategy and a new set of products coming to market in 2013. We continue to make the investments required to drive a larger % of our revenue in the high-growth adjacent markets: dense server, semi-custom, embedded, and ultra-low-power client markets. We are implementing three phases of our turnaround.
First, complete the restructuring of our business. This is a critical step in reducing our operating cost model to enable a return to profitability. Second, accelerate our business in 2013 by executing the delivery and launch of a new set of powerful product offerings. Third, transform AMD to take advantage of high-growth opportunities in adjacent markets where our IP provides a competitive advantage. As with any turnaround, it is critical we meet our commitments throughout each step. To that end, we effectively managed our cost, reduced our inventory, and maintained our cash above our minimum acceptable levels in the fourth quarter. Revenue of $1.55 billion met our guidance, while decreasing 9% sequentially and by 32% from a year ago. For the second year in a row, we had a successful Black Friday as sales of AMD-based notebooks in North America retail grew from a year ago.
As a result, nearly one in every three notebooks sold in U.S. retail in the fourth quarter were powered by AMD. We also saw a sequential increase in desktop microprocessor ASP based on strong channel adoption of our new Athlon FX CPU and A-Series APUs in the quarter. The value proposition of our latest processors drove demand and helped reduce channel inventory. Customers continue to embrace the differentiated value proposition of our APUs. We added Vizio as a new customer based on the industry-leading graphics performance and long battery life of AMD's APUs. Vizio's first AMD-based products will include two touch-based ultrathin and a Windows 8 tablet. Now turning to our server business. We recorded significant revenue growth for our SeaMicro dense servers, driven by large-scale cloud data center wins.
SeaMicro solutions are being deployed and evaluated by a broad variety of marquee customers, demonstrating the value of our investment in dense server and setting the stage for continued future growth. Our graphics business performed well in the quarter as well, highlighted by record workstation and gaming revenue and the launch of the Nintendo Wii U game system, powered by AMD graphics technology. To help accelerate desktop GPU sales in the channel and reinforce our long-term strategy to be a leader in gaming, we launched the Never Settle campaign in the quarter. This promotion was well-received by partners and consumers as we bundled some of the season's hottest games with our highest-performing graphics cards. In the channel, unit shipments and revenue for our high-end Radeon graphics grew, driving a richer mix and increased ASP from the previous quarter.
We will launch a follow-on promotion this quarter, pairing our highest-performing graphics cards with some of the most anticipated game titles of 2013. I want to return to the trends impacting AMD and the industry overall, and the steps we are taking to restructure, accelerate, and ultimately transform our business. We are well on our way toward implementing a new operating model that will allow us to achieve lower operating expense and enable a return to profitability. In addition, the new products that will accelerate our business in 2013 have passed several key milestones in the quarter as we began shipping Richland and secured solid design wins for the Brazos follow-on Kabini and our new ultra-low power Temash APU. Both Kabini and Temash are currently going through final internal validation in parallel with customer evaluations in anticipation of launches that are planned for the first half of this year.
AMD's first mobile GPUs, based on our Graphics Core Next architecture, also began shipping in the fourth quarter. Asus, Lenovo, and Samsung have already announced plans to offer these low-power GPUs, and we are seeing strong design win momentum for these GPUs because they deliver discrete-level graphics performance while stretching battery life in both traditional and ultrathin notebooks. Our longer-term transformation to drive growth in adjacent high-growth markets continues to gain momentum. In the server space, we'll combine our extensive 64-bit design experience, x86 processor IP, and ARM processor cores with our SeaMicro Freedom Fabric to continue to drive leadership as the industry transitions to dense servers. We believe we already have significantly more dense server customer installations than any other competitor, making our SeaMicro technology the most tried dense server solution available in the industry.
In ultra-low power clients, we bolstered our SoC design capabilities with new engineering talent, simplified our development processes to improve our execution speed and quality. Our Kabini and Temash SoCs for notebooks, tablets, and convertibles are exceeding our expectations. More importantly, the strong value proposition and differentiated features of these products are exceeding customer expectations as well. Kabini will be the first industry quad-core x86 SoC, and Temash is expected to be the highest performing tablet SoC available in the market, more than doubling the graphics performance of our current generation tablet offering. Finally, we have strong design wins for our embedded and semi-custom APUs. Initial products based on these APUs are expected to launch later this year, driving our embedded semi-custom business to more than 20% of our revenue mix by the fourth quarter.
In summary, we have made good progress in the fourth quarter with our turnaround. We delivered on our financial and customer commitment. Effectively managed cash, reduced inventory, and lowered operating expense in the face of a difficult market. We will complete the majority of our restructuring actions this quarter with the goal of reducing our overall cost structure by 25% from early 2012 levels by the third quarter of 2013. We are introducing strong new APU and graphics offerings in the first half of 2013 that will accelerate our AMD business, with the ultimate goal of returning AMD to profitability and positive free cash flow in the second half of this year. We are also making good progress on the multi-year transformation of AMD that will allow us to deliver 40%-50% of our future revenues from high-growth markets where our IP is a differentiator.
Our embedded semi-custom APU business remains on track to contribute approximately 20% of our overall revenue by the end of this year. Our SeaMicro dense cloud server business continues to gain momentum. We will remain focused on the steps we must take over the next several quarters to turn AMD around and deliver consistent profitability and growth. With that, I will now like to turn the call over to our newly appointed CFO, Devinder Kumar. Devinder?
Thank you, Rory. 2012 was a challenging year in which we made strategic investments and began to take the actions necessary as part of a corporate and financial reset to position the company for future success in a changing computing landscape. In the fourth quarter of 2012, we announced restructuring actions and operational efficiencies designed to reduce our expense structure and help return AMD to future profitability and free cash flow generation. The restructuring plan includes a reduction of AMD's global workforce in the fourth quarter of 2012 and the first quarter of 2013 by approximately 14% and will result in operational savings of approximately $190 million in 2013. By the third quarter of 2013, we expect to reduce operating expenses to $450 million per quarter, down 25% from the first quarter of 2012.
As we ended 2012, we also successfully amended our wafer supply agreement with GlobalFoundries to better align our wafer purchase commitments with current PC market dynamics, strengthen our balance sheet, and help us achieve our operating goals. Broader macroeconomic issues impacted consumer PC spend in the second half of 2012, as the challenges we faced in the second quarter continued through the end of the year. We faced a difficult selling environment, which negatively impacted our overall 2012 financial performance. Let me provide some details on the full year 2012 financial results. Revenue for the year was $5.4 billion, down 17% year-over-year. Non-GAAP gross margin was 41%. Non-GAAP operating expenses were $2.2 billion, lower than what we guided due to the restructuring actions and alignment of expenditures with business expectations. We achieved non-GAAP operating income of $45 million.
We managed capital expenditures down to $133 million in 2012, below guidance of approximately $200 million. Tax benefit was $34 million, better than guided, due to a one-time benefit of $36 million related to the SeaMicro acquisition. Cash balance as of the end of the fourth quarter was $1.2 billion, above the optimal zone of approximately $1.1 billion and well above the target minimum of $700 million needed operationally. Let's turn to the fourth quarter results. Revenue for the fourth quarter of 2012 was $1.16 billion, down 9% sequentially, driven by a revenue decline in computing solutions segment of 11% and a decline of 5% in the graphics product segment. Revenue declined due to lower volumes across both segments.
To derive non-GAAP gross margin and other non-GAAP financial measures, among other items, we excluded the impact of the lower of cost or market charge related to the GlobalFoundries take-or-pay obligation of $273 million, discussed in more detail in my CFO written commentary, which has been posted online. Non-GAAP gross margin was 39%, up eight percentage points quarter-over-quarter. Gross margin in the third quarter of 2012 was 31%. This was adversely impacted by an inventory write-down of approximately $100 million or eight percentage points. Fourth quarter 2012 gross margin also benefited from sales of higher priced second-generation desktop APUs. Non-GAAP operating expenses were $506 million, lower than guidance, primarily due to tight spending controls in the quarter. R&D expenses were $313 million, 27% of net revenue. SG&A expenses were $193 million, 17% of net revenue.
Non-GAAP operating loss was $55 million. Non-GAAP net loss was $102 million. Interest expense was $45 million, flat compared to the prior quarter. Tax provision for the quarter was $4 million compared to zero in the prior quarter. The non-GAAP loss per share was $0.14, calculated using 747 million basic shares. Adjusted EBITDA was $30 million, better by $65 million from the prior quarter's negative $35 million, excluding the LCM charge of $273 million. Switching to the business segments. Computing Solutions segment revenue was $829 million, down 11% sequentially. Client product revenue declined 13% sequentially, primarily driven by lower microprocessor unit shipments. Server business revenue increased sequentially, driven by an increase in unit shipments of compute systems focused on dense servers. Chipset revenue declined in line with microprocessor unit shipments.
Computing Solutions operating loss was $323 million, an increase of $209 million sequentially, primarily due to the $273 million LCM charge. Graphics segment revenue was $326 million, down 5% sequentially, primarily due to an 8% sequential decline in GPU revenue, partially offset by record game console revenue and record workstation graphic sales. Graphics segment operating income was $22 million, up $4 million from the prior quarter, primarily due to the higher game console revenue. Turning to the balance sheet. Our cash equivalents, and marketable securities balance, including long-term marketable securities at the end of the quarter was $1.2 billion, down $297 million compared to the end of the third quarter of 2012. The decrease is primarily due to operating cash flow of $286 million, which included a few specific cash payments as follows.
Payments to GlobalFoundries of $130 million, comprising of $50 million for the limited waiver of exclusivity and $80 million termination fee related to the take-or-pay agreement. Cash payments of $46 million related to restructuring actions taken in the fourth quarter of 2012. We exited the quarter above our target optimal cash level and well above the target minimum cash level of approximately $700 million. Debt as of the end of the quarter was unchanged at $2.04 billion. Accounts receivable at the end of the quarter was $630 million, down $53 million compared to the end of the third quarter of 2012, primarily due to lower revenue. Inventory was $562 million exiting the quarter, down $182 million or 25% from the prior quarter due to a decline in MPU inventory, primarily as a result of the amendment of the wafer supply agreement with GlobalFoundries. Turning to the outlook.
For the first quarter of 2013, AMD expects revenue to decrease 9% sequentially, ±3%. Gross margin is expected to be flat sequentially. Operating expenses are expected to be approximately $495 million, and we expect inventory to increase sequentially ahead of new product introductions and technology transitions. For 2013, we expect operating expenses to be at $450 million by the third quarter of this year, capital expenditures of approximately $150 million for the year, taxes of approximately $4 million per quarter. We plan to be free cash flow positive by the second half of 2013, and we expect to maintain cash balances in the optimal zone of $1.1 billion for the year, and well above the target minimum of $700 million.
Overall, we have taken significant steps to align our business with today's PC environment and position the company to capitalize on new opportunities in adjacent high-growth markets in 2013 and beyond. We expect our restructuring actions and our corporate reset to position us to return to profitability and free cash flow generation in the second half of 2013. I will now turn it back to Ruth for the Q&A. Ruth?
Thank you, Devinder. Operator, we're now happy to poll the audience for questions, please.
Thank you. If you have a question, please press star then one. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. If you have a question, please press star then one. Our first question comes from David Wong from Wells Fargo. Your line is open.
Thank you very much. Embedded revenues. Rory, I think you said you're still on track to hit 20% of revenues by 4Q for the embedded revenues. Can you give us some idea as to how things will ramp in the first, second, and third quarter? Will you get any significant embedded sales in those three quarters?
Yeah, David, we already have an interesting embedded business already in place. What we're going to see as we continue to invest in that area to leverage our differentiated IP and our leadership APUs, is a move into some of the higher growth segments. We've mentioned on previous calls that those are currently confidential, and they'll be announced as the year moves on. We should see us build that revenue starting in the next several quarters, and it ramps throughout the year to that 20% level.
Okay, great. My other question is, over the last several quarters, since you've been at the company and you've been forming your long-term strategic goals, has your thinking changed on how AMD should be transformed? Are there new segments you now think AMD should address or alternatively, areas that you've decided AMD should exit?
Yeah. It's interesting, David. Over the past 16 months of being here at AMD, it's clear that this company has a deep base of intellectual property. It has a huge asset pool in terms of the talented engineers and dedicated AMDers across the planet. This is an unprecedented period of change in the industry. Our strategy has remained basically fundamentally consistent throughout this. The only thing that's really changed is the time period that we need to execute that change and accelerate into that. First, it's a three-phase turnaround. One, reset and restructure the company. I think here in fourth quarter, you see us making that kind of progress, where we're delivering on that commitment, we're executing and managing our expense, we're managing our cash properly, and we've improved our inventory position significantly. With that, we're on a drive to get to $450 million by 3Q.
We need to execute and accelerate in 2013. That's where we hit the top of the curve of this restructure and reset phase and begin to move into the launch of these powerful new products that we have in place. At CES, Lisa and I spent a lot of time with customers, and they were excited about Richland, which is now launched, and they were excited about Kabini and Temash. Kabini, the follow-on to our most successful Brazos platform, and then Temash into the low power segment. These are very interesting products, and we're going to continue our industry leadership into the graphics segment. Dense Server had its best quarter ever and showing the foundation for future growth. Ultimately, where we have to take this company is we need to transform AMD to take advantage of where the market is going.
The high-growth segments, those areas where we can apply our innovation to lead. That's where we have to take the third part of our step, and we're making those investments now. In Dense Server, we're already seeing the early progress with SeaM icro that I mentioned. Moving into embedded semi-custom. This year will be 20%, but our ultimate objective within the next three years is to move to 40%-50% of our revenue in those segments. Obviously, the opportunity for us to build on our long heritage in the client segment around new form factors in the low power segment. These are areas where AMD can create leadership, diversify its portfolio, build on its long history of success and its differentiated IP for leadership. David, I think that's the one, two, three-step approach that we need to take.
Great. Thanks so much.
Thank you.
Our next question comes from John Pitzer from Credit Suisse. Your line is open.
Hi, this is Patrick Walsh. I am calling in for John Pitzer. I just had a quick question relative to some of the charges. In the fourth quarter, it looks like you guys took a larger than expected charge. How should we think about these kind of one-time charges going out through 2013?
I think if you look at Q4, you are right about the one-time charge, let me just take a moment to explain that. We had previously said a $165 million one-time charge and $110 million of our termination fee related to the WSA would be spread over Q4 and Q1. As we completed the analysis from an inventory standpoint, it was more appropriate to record almost the whole charge in Q4, we went ahead and did that is $273 million. At this point in time, even with the restructuring charge that has been taken in Q4, we do not anticipate anything significant happening in the 2013 timeframe. There are some actions that might be taken that would lead to it, at this point, I don't think anything is anticipated.
Okay, great. Thanks. Then just another question on the embedded segment. Just curious if you could give any color, how that mix could look over time in terms of what % of that is gaming versus other embedded revenue?
Patrick, this is Lisa Su. Just to comment on the embedded revenue. We put embedded and semi-custom designs in one bucket, and we would certainly see the embedded business tends to have a long design in cycle and then a long life cycle as well. We have very good visibility in terms of execution for the target that Rory mentioned, 20% of our revenue by the fourth quarter of this year. As we go forward, we would expect to add additional high volume design wins to that bucket. It's interesting, as we mentioned in the earlier comments, those parts are already in their validation phases here. We're working through those tracks, and that protects those key design wins that drive that 20% in terms of the revenue objective by 4Q.
We're on track and making solid progress in terms of the delivery of those key new solutions.
Great. Thank you very much.
Our next question comes from Hans Mosesmann from Raymond James. Your line is open.
Thanks. Hey, Rory, a couple of questions. You're going to end up being about six, maybe nine months ahead of the competition in a quad-core type of SOC. What are the skill sets that you're bringing forth to do that? The second question is, any commentary about the capacity additions that your rival is going to be making for next year and subsequent to that? Thanks.
Yeah, no problem, Hans. Hans, from a standpoint, we are working to create strength in our core businesses, at the same time working to capture those new opportunities outside of the traditional space. Over the next two years, you're going to see us rebalance that business, where we have a more even split across the non-PC portions and the PC portions. What's kind of interesting, as you talk about the leadership in that quad-core space of the SOC, that's a powerful product with Kabini that's launching this year. Again, it's in its final phases of test. I'm sure Lisa will comment in just a second. What we're trying to do in this segment is, again, to create the leadership that we've done in the past with Brazos, and then to apply that same kind of knowledge into Temash, which brings us down into the fanless tablet segment.
Now, from a talent standpoint, what we've really focused on is a mix of industry knowledge across the semiconductor space, and you've seen us bring in interesting players across the past 12, 15 months, players that are from the who's who of the industry. They know the semi-custom space. They know microprocessor design. They know semi-custom and, of course, embedded. These talents, also, you can see with the Kellers and Mark Papermaster and obviously Lisa Su, the experience to deliver that in industry-leading microprocessor design. That's why I think you're seeing some of that progress around Kabini. Lisa, you want to add a thought or two?
Yeah, Hans, just to answer the question about the tablet category, we're very excited about our Temash tablet because it really is satisfying a new space in terms of performance tablets that's separate from what today's consumption and sort of the higher power tablets that are out there. I think it is an opportunity for us to lead. I think from a mobile standpoint, this is where we think there's a sweet spot for full Windows 8 capable tablets and extending from down to quad-core to very low power points. As we build out this part of the roadmap, it's about system on chip designs, getting much more flexible in how we reuse IP, and getting products out to market at the right time. Having these products out the first half of 2013, very important to catch the strong back-to-school cycle.
Thanks. The follow-up was that comment on capacity.
Hans, in terms of capacity, what were you curious about?
Yeah, that your competition across the way there is going to add a lot of capacity, it seems.
Our focus, again, I think, is clear and very focused. We've got to stay on this reset restructure, get that done and behind us. I think we're moving through that well in terms of getting the cost model, then getting into the execution of the acceleration. That's the key. It's about getting those products in market. Lisa just talked about it in terms of Kabini and talked about it in terms of Temash. It's around that embedded segment. That's the key for us, and the customer acceptance, whether it's in the industrial segment, the gaming segment, our traditional OEMs or ODMs, they are very interested in the products that we're creating in 2013. That's the key for us. We do that, we correct the turnaround, we get to the lower operating model, we return to profitability and growth.
At the same time, Hans, continue to focus on where the market is going around dense server and around the areas of that tablet mobility segment, as well as continue to expand into the embedded semi-custom opportunity. I think we have to focus on what we're good at and getting the products executed. That's key for us.
Thank you.
Thanks, Hans.
Our next question comes from C.J. Muse from Barclays. Your line is open.
Yeah, good afternoon. Thank you for taking my question. I guess first question on gross margin guide of flat despite revenues down 9%. Can you walk through the drivers there between mix, any sort of moving parts with the WSA, with GF, or any other drivers that we should be thinking about, and then how we should think about the trajectory post Q1?
I think all the factors you mentioned come into play from a product mix standpoint. The WSA is behind us. We renegotiated that and amended that as we announced in December, so it is in effect for 2013. From that standpoint, we are projecting in Q1 of 2013 that we'll be able to maintain approximately flat gross margin at 39%.
Okay, that's helpful. I guess given what you've seen and what you're guiding to here on the gross margin side, is there an update in terms of what you expect breakeven-wise in terms of top line on an operating level basis? I believe a few months ago you had said $1.3 billion.
I think-
Sorry, go ahead.
Yeah, rather than focus on that, I would just like to reiterate and maybe add some color on the guidance we're providing for 2013. The PC market, as Rory said earlier, we see it as challenging for the next couple of quarters at least, therefore, we're not providing any guidance from a revenue standpoint for the year. From an OpEx standpoint, we have a trajectory to get to $450 million by Q3 of 2013. CapEx, as I said earlier in my remarks, $150 million for the year. Finally, free cash flow positive and profitable by the second half of 2013. That's the guidance we are providing for the 2013, from a viewpoint of the financials, as we sit here right now in the early part of 2013.
That's helpful. Last question from me. In terms of the cash outflows in Q1, can you confirm what those positions are? I believe you owe $175 million for SeaMicro and another $175 to GF. Is that correct? Are there other moving parts there?
You are correct. Luckily just for one of the $175 that you mentioned, the SeaMicro acquisition happened early part of 2012. We paid for that. That was all done as part of last year. You are correct about the $175 to GF to be paid in Q1 of 2013. The other payment that we have, which I mentioned in my remarks, is the $31 million for the severance charges for the actions we are taking from a restructuring standpoint. Those are the two specific payments that will occur in Q1 2013. From our standpoint, as I've said several times, from a cash standpoint, we have a target minimum of $700, but we plan to maintain a cushion between the $700 and actually maintain closer to the $1.1 billion, which is what we call the optimal cash balance.
Thank you.
Our next question comes from Joseph Moore from Morgan Stanley. Your line is open.
Great, thanks. First, just a quick follow-up on the last question. Where do you stand on the sale leaseback of the headquarters?
It's in progress. We are targeting to complete that within this quarter. We should net somewhere between $150 and $200 for the sale leaseback transaction. The only clarification I'll make there is this is the Austin campus. The headquarters are still in Sunnyvale, California.
Okay, great. Thank you. With the R&D, you've taken a lot out of R&D spending already. You're going to take that down further by Q3, but it seems like there's more on your plate now with the ARM server roadmap, SeaMicro, the embedded opportunities. Is there something coming out of R&D to let you make those cuts and feel like you're not cutting back, or is there a risk that you cut too far in any one area?
No. The focus around expense management in terms of driving that OpEx line is really across all parts of the business. We've tried to focus those areas where we thought there was efficiency and productivity to be gained. For example, there's opportunities for us to dramatically reduce the number of different process technologies we're running in our various foundries. That drives all the IP to be written to multiple libraries. We very much focused in terms of efficiency to drive the number of metal layers, the way we reuse our IP, the IP development process that Mark Papermaster and Lisa Su have implemented through our technology board. We are seeing efficiencies and productivities through that. At the same time, we're driving to improve our focus in terms of the types of design wins we go after.
It's really not the best business practice to go after just a gigantic number of design wins. We want those high runners, those design wins that drive volume. Every design win drives cost, drives back-end execution, testing resources. When we drive to hundreds of design wins, some of those are 25,000, 50,000 units. Those are not efficient. We've been focusing with Lisa's business unit teams and John Byrne's sales team to really identify those design wins that get the volume and get to the most efficient usage. That's how we're driving that kind of focus in terms of efficiency and productivity, and those are just a few. We've implemented the financial transformation system on the back end. John Dougherty on the supply chain has an intense focus across each of his processes to drive that efficiency, to improve our execution.
What's interesting, while we had some schedule challenges in the middle of summer last year, the second half of the year, the progress that the business unit teams and the technology teams have made on our 2013 roadmap part has been very consistent and on schedule. I think we're making the right moves to push for that efficiency, look for the reuse of IP, create data ambidextrous architecture that's reusable, to drive efficiencies across the portfolio. In fact, we announced a new Senior Vice President, Chekib Akrout, who is driving that transformation and efficiency focus across our entire business as a full-time job.
If I may add, I'll just add one comment to that. If you look at our OpEx in Q1 2012, it was closer to $600, and in Q4, we've ended closer to the $500 level. We've already taken out approximately about $100 million of expenses. The restructuring actions that we took from headcount and some facilities actions that we are putting in place happened very late in Q4, and obviously we have some going on in Q1 2013. Those benefits, obviously, as you know, lag the actions, and we'll see the full benefit of that expense deduction in Q2 of 2013, and that should help us, while protecting the R&D investment that Rory talked about, to get to the $450 by Q3 of 2013.
Again, at the end of the day, our company hinges on the products that we create. 2013 roadmap looks strong. It's across the board in terms of the execution, the schedules, and the launches are tracking right to the schedules that we laid out. We're working on 2014, 2015, and 2016 in terms of how to transform our business, in terms of those high-growth segments, and how do we continue to leverage our leadership in the client space and the graphics space to continue to deliver that product focus. We have protected through this process of budgeting those areas where we know the highest growth will come over the next several years.
Great. Thank you very much.
Thanks, Joe.
Our next question comes from JoAnne Feeney from Longbow Research. Your line is open.
Hi. Thank you. I wanted to go back to the discussion about gross margin. In the last quarter, desktop was stronger. I was wondering, first of all, if you could clarify whether that was purely a mix shift within desktop, say, on lower volume shipments, or if you actually saw an increase in units sequentially in the fourth quarter, in addition to, it sounds like, an ASP rise. That's sort of number 1. Number 2, what are you thinking about for mix for 2013? Do you see more lower end in PCs and tablets? How are your margins across those different kinds of products? What do you think could drive either gross margin to stay where it is or perhaps to expand in 2013?
Okay, JoAnne, this is Lisa. Maybe let me take some of the comments on sort of the mix on the gross margin in the fourth quarter. In the fourth quarter, I think, we took a very balanced approach to managing the margins. I think we saw some positives and some negatives in the desktop business. We did see an increase in desktop ASP, primarily because we introduced our higher-end Athlon FX series, as well as the new A-series Trinity APUs into the channel, and that drove ASPs up a little bit on the desktop side. There were some competitive pressures on the notebook side. All in all, very balanced approach to the margins. I think as we go forward for 2013 mix, we're not talking about full-year gross margins. Devinder mentioned our guidance into Q1.
When you look at the mix there, we would expect that as we get into the second half of the year, we will have the higher end of the A-series, the A-8s and A-10s, but there's, of course, a balance with what happens in the market. All in all, I think that's where we are.
Okay, a quick follow-up. How much are you, let me back up. The changes that you have been talking about, you and Rory Read and Mark Papermaster, on simplifying the manufacturing process, reducing the number of wafer processing steps, for example, adopting a more standard process, presumably is expected to do two things: one, lower unit costs, and two, speed your time to market. I guess from our perspective of trying to understand where gross margin might go, can you give us any sense of the quantitative impact that you foresee from those simplifications on the cost structure?
From a cost structure, that's part of the way how we drive to the 450. That's one piece of it, as I kind of highlighted, JoAnne Feeney. The idea across the development process, the way we interlock now, the focus that we have. In the past, custom, hugely long, complex designs were the way the PC market moved. In the future, it's going to be quicker, low power, more efficient design. What also you get from this approach, obviously you'll get time to market, you get lower cost, you also get better yield, which is also important in terms of really managing the business. It has benefits across the board. Speed of execution as we move forward in a model where you're going to see a consortium-based model and the proprietary control points of some of x86 break down, these kinds of models are going to be key.
We're going to stay the course on that. That's helping us on the 450. From the gross margin, I think we've given the guidance in fourth quarter. We need to focus on executing first quarter as we've given the guidance, excuse me, on first quarter. From there, we'll go forward.
Okay, just one quick follow-up. It sounds like the simplification of manufacturing is doing both. It's reducing OpEx and potentially improving your gross margin. Perhaps you could just give us an update on how the spread is on gross margin across your different products, from your mainstream Trinity Richland to your lower-end Temash and Kabini. Is that something that's widened or narrowed, or can you really not talk about that yet?
JoAnne, we wouldn't get into that level of detail. I think we've given a good outlook in terms of the guidance for 1Q.
Okay, thanks.
Thank you, JoAnne.
Our next question comes from Glen Yeung from Citi. Your line is open.
Hey, Glen, are you there?
Hi, this is Adeline. Yes. Hi, can you hear me?
We can. Go right ahead. You're not Glen. I know Glen's voice.
Yes. It's Adeline for Glen Yeung. Can you discuss your gaming consoles, which is a potential market for you, and, we just want to know, when can we expect revenue to ramp, and is that part of the inventory build that you've discussed in your CFO remarks? I have a follow-up.
Devinder, why don't you touch on inventory, how we focus and manage it, and how it progresses from here, and then we'll touch on our embedded strategy, which we've covered before for Adeline.
Yes, I'll do that. Hey, Adeline. Inventory, as you observe, down 25% quarter-on-quarter, but that's largely in the MPU inventory space and partly and largely due to the WSA amendment. From my standpoint, if you talk about inventory going forward, it's more in the $650-$700 range, is what I would call it, for 2013. In particular, as you can imagine, as new business opportunities arise, new product introductions take place, and in particular, when you have process technology transition, that is going to lead to some increase in inventory, and that's where it will be. I'll let Lisa or Rory answer the other part of the question.
Adeline, on the gaming revenue, I think we talked about the Nintendo Wii U that was announced in the fourth quarter, and that was part of the revenue that was reported in the graphics segment. In terms of going forward, what we've said is, the embedded and the semi-custom business will ramp over this period in 2013. We are on track for that 20% revenue target by the fourth quarter.
That would be more second half than first half. Can I make that assumption?
Our embedded business will continue to grow toward the 20% in the second half.
Okay. The follow-up is that now that you are seeing some success in the dense servers market, do you have an opinion about ARM versus x86? Did that strength at SeaMicro that you saw came with Intel or AMD processors?
The dense server market is where we believe the growth is in the server business, so we were pleased with the progress of our SeaMicro Systems in the fourth quarter. I will say that in the fourth quarter, we started shipments of AMD-based SeaMicro Systems as well as Intel-based, and they've gotten a good reaction. I think as we go forward, we are committed to both ARM and x86 in the appropriate markets. We've stated that we will be doing ARM-based server chips, and that will be in the 2014 timeframe. We continue to offer a strong lineup of x86 as we go through both our client and our dense server business.
This is all consistent with our focus to drive reusable IP, an SoC, true SoC methodology, and an ambidextrous architecture. This is an area where AMD can clearly differentiate and create leadership in the marketplace.
Great. Thank you.
Our next question comes from Jim Covello from Goldman Sachs. Your line is open.
Great. Thanks very much for taking the question. This is Mark Delaney calling on behalf of Jim Covello. I was hoping first you guys could talk a little bit about your guidance on the revenue for the first quarter and why you think it is that you're guiding below your biggest competitor.
From my perspective, as I look at the PC market, Mark, that market's going to continue to be choppy in 2013, particularly in the first half. Remember, if you go back to 1Q last year, there was a lot of concern about the flood that affected Thailand and the hard disk drive, and it was quite surprising, as we kind of suggested, the supply chain would be quite resilient. The numbers were quite strong out of 1Q. We think there'll be continued chop and pressure in that first half of 2013, and this market's a bit dynamic right now. We do think Windows 8 is a very important event in the industry, and I think that impact or effect will build over the course of the year. We expect the second half to be stronger than the first half, from my perspective.
If I looked at the overall year, weaker in the first half, stronger in the second half, probably a net flat to slightly down. That's our view of it in terms of where we're expecting the market and how we're planning to balance that in terms of the gross margin and in terms of the revenue line. That's how we came to the conclusion.
That's helpful. Thank you. As a follow-up question, the mobile space in particular and tablets, there's been a lot of new entrants, companies that don't historically participate in the PC market that have been introducing products. Now that you guys are closer and farther along with some of your products for that space, I was hoping you could help us understand to what extent you guys are working closely and partnering with some of these new companies that you traditionally haven't worked as closely with.
Well, I think what's important is that we see that same high growth opportunity in those new form factors. You can see it with the introduction of our relationship with Vizio just here at CES, and that introduction of those two ultra-thins and obviously in the tablet space. You're going to see us continuing to work across OEMs, in the embedded space, new partners, and across the ODM segment to get that expansion and begin to build that market. Again, our focus is to build that over a two to three-year period to drive that to 40%-50% of our revenue across those high-growth segments. Lisa, any additional comments you'd like to add?
No, I would agree. I think that the key with the mobile space is that there are a lot of new innovations happening, both in the traditional OEMs as well as in some of the new entrants, as you mentioned. Yes, we are working across the board with a number of customers with these new products.
Thank you very much.
Thanks, Mark.
Our next question comes from Steven Eliscu from UBS. Your line is open.
Yes, thank you. Actually, regarding the last question, I want to take a glass-half-full approach. This quarter with your guidance, it appears you're going to lose a lot less market share to your competitor. Is this about Richland ramping, or is this just about clearing excess inventories that you had on Llano in the channel last year?
I think we've had a clear focus for the past several quarters to improve our channel execution from a push model to a velocity model where I have sell-through tracking and the relationship and data all the way through the channel. I think we're making steady progress in Q3, Q4, as we move forward there. I think that inventory and our finished goods position, Devinder touched on it with our WSA work and the focus. We, at the end of Q2 and even into Q3, were seeing pressure in terms of potentially building more inventory. While we did have to take a charge in Q3 to address some of that, I think we've done a strong job in terms of focusing on that inventory. That positions us for the new product introductions that we talked about at CES, Richland, Kabini, and Temash.
I think that's what positions us to move forward and begin to build this turn and acceleration in 2013 with the objective to get to profitability in the second half and positive free cash flow at that time frame again. That's the key to our business.
Okay. That's helpful. Thank you. As a follow-up, you lost a lot of graphics share in Q2 and Q3, would want to get your view if the new GCN parts reverse that share materially, where you get back to a level you were in earlier in 2012, or is this just about stabilizing share since your competitor isn't standing still either?
Yeah, Steve, let me take that on the graphics business. Look, the graphics business is very strategic to our overall strategy. I will say that we have lost some share over the last couple of quarters. I do believe we're at the low point in our graphics share based on what we can see in terms of new design activity. We have a very strong channel approach with our Never Settle bundle that we talked about last quarter, and we will be building upon that this quarter. We also launched our new HD 8000 series at CES that's now shipping in a number of mobile OEMs. I think the share is indicative of decisions that were made 12 months ago. What we're looking at is really a strong portfolio as we go forward in graphics.
I'll make just a brief comment from the standpoint of my experience as I joined AMD. The graphics team here that we have is leadership. The kind of IP, the kind of engineering skill, the ability for them to innovate and drive. I don't know if you had a chance to walk through, Steven, the CES work that we showed there.
Yes.
I thought you saw some really interesting application of how you could create a surround compute environment leveraging the graphics capability. Imagine that kind of solution moving forward. Say they're building a brand-new building and that building's not even built. You could create with our technology, a sales room that created the windows and the graphics representation and say, "Well, what's this building on the 22nd floor look like at night?" All of a sudden, the graphic engines fire up the windows. You can see that world, and you could apply that kind of surround compute that we see emerging, this new trend of convergence that the cloud is fundamentally driving, and it's our graphics IP that's going to be a fundamental enabler of that future. That's the world we're chasing. We're not going after the past.
That standalone data and application on a singular device protected by a couple of proprietary control points, that era is ending. The era in front of us is an interconnected world, where there's almost a tsunami of new devices that begin to emerge, and they're all graphically linked because the data and information, the applications are running on cloud dense servers, and these mega data centers. They're distributing information across the whole set of client devices. They're efficient, they're low power, they're across embedded devices. This is the opportunity, Steve, in front of us to really get after it and where we can see graphics taking the business moving forward.
Mm-hmm. Thank you.
Operator, we'll take two more callers, please.
Our next question comes from Romit Shah from Nomura. Your line is open.
Romit, are you there? Are you out there?
Hi, Rory, t his is Sanjay for Romit Shah.
Oh, excellent. Perfect. I thought you had.
Question I wanted to ask is if you could share some updates on 28 nanometer progress, at GlobalFoundries as it relates to some of the issues you faced last year at 32 nanometer.
From our perspective, our supply chain has focused with our foundry partners, across 2012 in improving execution. As I've commented several times before, the work that we've done with GlobalFoundries and across the foundry supply chain environment has significantly improved with that kind of a disciplined approach. We introduced last year the first 28 nanometer graphics products, and leadership, and now as we move forward, we're positioning ourselves with that refocused, better executing supply chain to move forward in 28.
Yeah, Sanjay, just to your question about 28 nanometer and GlobalFoundries, we're pleased with our overall 28 nanometer bring up in all of our foundries. From a 28 nanometer standpoint, we feel very good about where the technology is and what it's delivering for us from a product standpoint.
One of the things, Sanjay, that we're focused on is to make sure that those parts are delivered on time to the market, the PC launches, and that we get the parts in volume as we ramp through it so that it's not just an announce and a back end. These are focused, in terms of the execution of the supply chain to build through the ramp at the beginning of the launch. That's something we've focused very hard not to repeat the Llano kind of events of 2011.
Okay, thanks. As a follow-up, I wanted to go back to the gross margin question, and I was wondering if you could give us a sense on your embedded and semi-custom business. What's the gross margin of that product line relative to your PC business? Is it lower? Is it higher? If you could give us any color there.
No, we don't get to that level of granularity, Sanjay.
All right. Thank you so much.
Our last question comes from Vivek Arya from Bank of America. Your line is open.
Hi, this is Ashish Rao for Vivek Arya. Thanks for squeezing us in. Question for Lisa. Could you provide an update on when you expect Piledriver architecture Opteron servers to come out? Related to that is, could this refresh your server business, or do you think long-term growth is more a function of SeaMicro dense servers as well as the ARM-based servers?
Yeah. When we talk about our server business, we are definitely focused on increasing our investments in the dense server portion of the business. We recently announced our Abu Dhabi lineup in Q4. We also announced earlier this month an Open Compute reference platform, that was based on a system that we call Roadrunner. Those are both based on the Piledriver architecture. As we go forward, as we said, we are going to ensure that we double down in the dense server business. That includes our SeaMicro fabric as well as our chip technologies with APUs and CPUs.
Okay, cool. Devinder, let me try asking you the semi-custom embedded business question in a different way. You have noted that you expect it to represent about 20% of sales exiting the year. Directionally, how should we think about gross and op margins, both on a % as well as on a dollars basis, as you go from progress through the year?
Yeah, I think, as I said earlier, the PC environment, it continues to be dynamic. Our business model, based on what you observe with the 20%, obviously, is evolving. We are providing guidance for gross margin for Q1 2013. We're not providing guidance beyond that time period.
Okay, thanks.
Operator, that concludes today's earnings conference call. We'd like to thank everybody for participating and for you to sign off now, please, operator. Thank you.