This has been an interesting quarter for Advanced Micro Devices (AMD). As an AMD long, I was quite happy with the previous earnings report.
When I look at my AMD position, the things that concern me the most are:
During the previous earnings call, none of my personal "trip wires" was triggered to give me cause for great concern, but the stock sold off sharply on what I felt was a positive report nonetheless.
In this article I would like to quickly touch on my concerns regarding AMD, look at each of its operating segments, and end with a conclusion as to why I feel positioning for Cloud growth and "console business earnings power" will likely be the focal points. Each heading will contain a short summary at the end to provide users a "condensed" version of that paragraph.
Points 1 and 3 go hand in hand, as some of the biggest penalties AMD has faced in recent times come from being unable to meet contractual obligations with Global Foundries.
During a previous article, I covered this in great detail. If you have not already done so, I recommend reading the paragraph (linked to here) that begins with, "This is the most important paragraph." This article also contains cut and pastes from previous financial statements from AMD, so it presents a source of consolidated information.
During the Q2 conference call, Mr. Kumar stated that about half of the contractual obligations had been met thus far for the year. During this call I am looking for an answer as to how much more obligation AMD has with Global Foundries during the 2013 FY. Another important point regarding the WSA (Wafer Supply Agreement) between AMD and Global Foundries is the details for 2014 pricing have yet to be announced, and are to be negotiated during 2013. News regarding 2014 pricing could be announced either tomorrow or shortly thereafter.
Regarding point 2, specifically for CPUs, reports and data point to AMD gaining slight marketshare, or at a minimum maintaining market share, against rival chip maker Intel in the consumer PC space. Based on the timing of Intel's Bay Trail launch, AMD has had a full quarter essentially with no competition from Intel on the lower power chip front. Note this dynamic will change during Q4 as products based on Bay Trail come to market, and Bay Trail traction will be worth paying attention to. The following graphic is taken from AMD's Investor presentation from Q2, and is based on research data from Mercury research:
I have explained before that I do not feel the power consumption characteristics for Bay Trail will be damning for AMD in the form factors where AMD is competing (I, II). While performing comparable workloads, Intel does have a performance/watt advantage to AMD, but Bay Trail is paired with a weak GPU. Consumers that would like to use small form-factor notebooks that are interested in any thing that uses the GPU heavily will be better served by a Kabini offering than a comparable Intel chip.
Regarding GPUs, a report by Jon Peddie Research points to AMD gaining traction in the discrete GPU market against Nvidia, and this market share was gained with an aging line of GPUs.
*Summary* As of the Q2 earnings report, AMD had extinguished half of its requirement for the year with Global Foundries, and has a competitive portfolio of products. Regarding debt, AMD needs to become cash flow positive soon in order meet debt obligations. AMD is expected to post a positive EPS this quarter, and this positive EPS is a starting argument for many longs. If AMD misses expectations, I expect to see a sell-off unless there is other good news to offset the miss.
AMD's major reportable operating segments are Computing Solutions and Graphics and Visuals.
Regarding Computing Solutions, AMD drove a positive surprise last quarter partially due to launches of its new low power Jaguar based chips, which I explained above I feel will remain competitive in small form factor notebooks and larger devices. During Q2 and Q3, there has been a steady roll-out of Kabini and Temash based devices from most of the major PC OEMs. Note that not all of PCs linked to on the Newegg website are based on the Jaguar cores, but it shows that Toshiba, Acer, Lenovo, Gateway, and HP are all featuring PCs using AMD hardware. Interestingly, reports also suggest that Dell will be launching a Jaguar based PC from AMD.
AMD also announced the quad-core A4-1350 Temash processor. I never had extremely high hopes for the dual core Temash part. On a performance basis, it would be competitive with Intel's out-going line of Atom processors. But based on reading through product reviews of Windows 8-based Z2760 (older) Atoms, the performance seemed to leave many consumers wanting. Some users only care about price, and dual core parts will serve Internet and media consumption needs of these users at low price points, but I do not think this market segment is that large given the large steps up in performance that can be obtained for minimal extra cash.
The A4-1350 is marketed using numbers that more closely resemble Intel's "scenario design power" power marketing numbers. The A4-1350 offers a quad-core part that operates in a 3W 'SDP' power envelope. I will not be surprised to finally see a tablet or two released with this part, as it brings quad-core performance to the table.
Regarding the overall PC market, IDC numbers show slight sequential growth from Q2 to Q3. Intel stated during its Q3 conference call the company could see signs of the PC market beginning to "bottom out," and saw PC CPU sales rise 4% sequentially. Given that IDC numbers point to a ~6M unit increase sequentially in PC shipments, and Intel has somewhere around 80% of the PC market share (~85%, but adjusted to account for data from Mercury research to make my guess more conservative), Intel likely saw ~3M extra PC CPU sales to PC OEMs (80% * 75M (~IDC Data from Q2 shipments) * 1.04 (increase in shipments as stated by Intel) = ~63M chips sold by Intel to PC OEMs). This would leave AMD with the remaining ~19M chip sales (81.6M per IDC numbers - my estimation for Intel's CPU chip sales in the most recent quarter). 19M chip sales to OEMs as compared with a somewhere around ~16M likely during Q2 represents a potential ~20% increase in CPU sales for AMD QoQ, and this is only accounting for PC OEMs, not chips purchased through retailers such as Newegg.
Unfortunately there is not enough resolution in the data to make accurate predictions and calculations, but the numbers do tell me there is a likely chance that higher PC CPU chip sales could give Computing Solutions revenue a boost.
Lastly, in my previous article outlining AMD as a potential cloud/microserver play going forward, I noted that AMD will now be offering Opteron versions of the SeaMicro servers. The SeaMicro server acquisition gives AMD the ability to sell directly to consumers, bypassing some of the hardware vendors. Depending on Verizon's purchase timeframe for the SeaMicro servers, this could impact Q3 financials, if the sales were not baked into Q2. The most detailed information I can find regarding the deal states that Verizon aims to start bringing these servers online during Q4 this year.
Another smaller, but potential positive factor in Q3 financials is the roll-out of Kaveri based motherboards. During Q2, chipset revenue was described as "flat sequentially." While not a major revenue driver, it could add a slight increase in revenue (think in orders of magnitude in the $10 million range).
I feel AMD has a competitive portfolio of products, especially among the client space. IDC numbers, combined with industry estimations of respective market shares, point to a potential for AMD to surprise with higher than expected PC division sales. Given that AMD stated during Q2 it expected PC sales to pick up in the back half of the year, it is hard to know how much of this rise was already baked into management's guidance. The SeaMicro acquisition and introduction of Opteron powered SeaMicro servers give AMD another potential avenue for higher revenues in the Computing Solutions group moving forward.
Graphics and Visual Solutions
This section will be shorter and will not be summarized.
The first stated reason for Graphics and Visual Solutions posting a break-even quarter during Q2, compared to $16M in profits during Q1, was lower console royalties. This should not come as a surprise as consumers are likely busy pre-ordering the next generation consoles. Both Sony and Microsoft have stated they expect record breaking sales for their new consoles. These record breaking sales should serve to offset the decline from Q2 in console royalties.
AMD has also refreshed its GPU lineup, and its "newest" products appear very competitive across all price points. I point out the word "newest" is because the products are by and large re-branded versions of its 7000 series GPUs. The only new GPU will be the Hawaii, but at this point I believe it is debuting too late to make much of an impact on Q3, and will likely be more of a factor in Q4. Note the Hawaii GPU is shaping up to be an excellent product based on leaked reviews, and looks to compete with Nvidia's $1000 Titan.
AMD also revamped its "Never Settle" campaign, removing two of the newer AAA titles (BioShock Infinite and Crysis 3), and reduced the number of games a user can receive across the various tiers of the program, while allowing users to pick and choose their free titles. During Q2, AMD stated it had lower ASPs for its GPU lines. Given the new products are typically price reduced from their 7000 series counterparts, I will not be surprised to see lower ASPs again this quarter across the GPU lines. Part of the lower ASPs could be offset by the changes in the Never Settle programs however.
Lastly, AMD has had record revenues in the previous few quarters in the workstation and server class GPU cards. Apple will be launching its refreshed MacPro update this fall, which will use more expensive workstation class cards based on AMD GPUs. This design win could continue to fuel record professional GPU revenues, based on how much inventory Apple builds for the launch, and the timing of the shipments from AMD to Apple.
Conclusion and Things To Watch For
During the previous earnings report, analyst questions regarding console margins were the focal point. Given this quarter we will finally see the direct impact of console silicon sales on AMD's financials, I suspect this again will be the focus. If AMD can silence arguments regarding the earnings power of consoles, I will view this as a huge positive for the share price; it would take some of the wind from the sails of the console margins argument.
Given the Verizon deal with AMD, I also expect a large focus to be on this going forward. The earnings call could shed some light on when we can expect to see the impact of server sales on AMD's financials, and possibly the magnitude.
My biggest concern remains contractual agreements with Global Foundries, so this will be my main concern going forward. I would also like to hear some insight as to how the Kaveri ramp is going. I expect to see AMD's new big-core CPU architecture to be a substantial revenue driver during 2014. Rumors I have read point to this APU being manufactured at Global Foundries. I would also like to see an update on where AMD stands with wafer purchase requirements.
Given AMD's strong second quarter in Computing Solutions, I expect sequentially improving PC sales to be a positive factor going into tomorrow. I think Graphical and Visual solutions revenues (excluding consoles) will be fairly uneventful. The focus in this operating segment will be on console earnings power, and if AMD can manage to silence the critics, I will look for the stock to trend higher, barring any other bad news is announced and Washington does not overshadow the conference call.
Additional disclosure: I own both options and stock in AMD, and actively trade my position. I may add or liquidate shares/calls at any time. I am also short NVDA through a very small number of puts. I may liquidate the NVDA puts within the next 72 hours.