On Wednesday AMD’s CFO Devinder Kumar spoke at the Credit Suisse Annual Technology Conference in Scottsdale. Mr. Kumar talked about the transformation of AMD’s business model and recent reorganization. In his comments he covered a few things that we found interesting.
As it stands today 35 to 40 percent of AMD revenue is coming from its Embedded, Enterprise, and Semi-Custom business. This is a business that is expected to continue growing over the next three years thanks to new semi-custom deals, the release of AMD’s ARM server chips, and the reuse of AMD’s core IP in the embedded market.
Touching on AMD’s semicustom business Mr. Kumar was questioned on whether or not AMD believes that 2014 would be the height of its revenues from the Sony and Microsoft console deals. In past console release cycles the first full year of production tends to have the largest revenue. For AMD that first full year is 2014. Somewhat surprisingly AMD is not convinced that 2014 will be the highest year for revenue from the current generation of consoles. According to him 2015 will could be the biggest year for console revenues and unit shipments because of the continued replacement of prior generation devices.
More importantly though AMD is seeing higher margins from the console deals than they expected. Because AMD is producing console chips for all three major gaming platforms they have full visibility of how many total units are going to ship in that market. This allows AMD to extract more efficiencies from their manufacturing, testing, and packing processes than they would be able to otherwise.
Originally AMD expected semi-custom operating margins to be in the low teens. For the initial year and a half of production actual margins have been in the mid-teens and over the past few quarters those margins have been consistently rising. Mr. Kumar attributes this increase to cooperation with AMD’s foundry partners which means that these gains are likely a combination of higher production yields and reduced inventory thanks to AMD’s complete view of the market.
The icing on cake though is AMD’s belief that as the cycle continues they will be about to achieve operating margins north of 20 percent on the semicustom business. This is a pretty significant increase in profitability from the low teens operating margins that AMD original billed the semi-custom business for.
Another aspect of AMD’s semicustom business is its recently announced design wins that will arrive in 2016. One win is for an ARM-based product and the other is for a traditional x86 product. Revenue is expected to total about a billion dollars over the course of three years. While Mr. Kumar was happy to mention those deals he was lax to give any more specific details.
Moving to the embedded business Mr. Kumar noted that in total it’s a 9 billion dollar business on a yearly basis of which AMD currently only addresses two billion dollars of. Of that two billion AMD holds between eight and ten percent market share. They’ve also seen consistent year over year growth in this business since 2013. AMD saw double-digit growth in this business from Q2 to Q3 in 2014 and are expecting to see double digital growth year over year for 2014. The margins from sales in this business are significantly higher than AMD’s corporate average. Thus continued market share growth in this segment will boost AMD’s overall profitability.
Mr. Kumar was quick to keep expectations in check though saying that the revenue contribution to AMD’s bottom line was in the tens of millions as opposed to the hundreds of million like the semi-custom business. This small revenue base is made up for by the long design life cycles in the embedded space which provide very consistent revenue. From these comments it appears that AMD expects the embedded segment to be a consistent, if small, profit center for them and are investing in it accordingly.
Moving to AMD’s foundry partners AMD’s 2014 wafer supply agreement (WSA) with Global Foundries specifies 1.2 billion dollars in purchases. Mr. Kumar did not comment on whether or not AMD expected to meet that goal unlike prior years. But he did spend quite a bit of time talking about AMD’s relationship with Global Foundries. The two companies are currently in talks about their 2015 wafer supply agreement and Mr. Kumar was very positive about the current state of that relationship. He went as far as to declare that the current relationship between the two companies is the best it has ever been in history and that he is very pleased with their performance. He also praised Global Foundries in its recent turnaround in execution and the FinFET coöperation deal with Samsung.
When pressed on AMD’s outlook towards advanced process nodes Mr. Kumar restated the same sentiment that we’ve heard from AMD’s last two CEOs: AMD does not need to be on the bleeding edge of process technologies to remain competitive. While there are some that may take issue with that statement it’s very clear where AMD stands on the issue.
As far as AMD process technology roadmap goes most of its products today are produced on the 28nm process node at both TSMC and Global Foundries. Looking forward Mr. Kumar said that certain products will be moved onto the 20nm node and then AMD will move to FinFETs from there. He also noted that AMD is now producing computing, graphics, and semi-custom products all at Global Foundries which is a big change from 2012 when at least two of those product lines were produced only at TSMC.
Touching on the weakness in AMD’s computing and graphics business Mr. Kumar cited AMD’s conscious choice to avoid low volume product wins at the cost of market share as the main driver behind falling revenues in this segment. He also cited AMD’s reliance on the declining consumer and channel segments of the PC business, as well as tougher competition from Intel, as additional reasons for their shrinking revenue numbers.
On brighter note Mr. Kumar described the recent restricting of AMD’s computing and graphics group as the key to increasing the efficiency of this business and said that loses from this segment had declined from 90 million dollars to 20 million dollars year over year. AMD sees the same customers for both its graphics and computing products which is why the creation of the group made sense to them. Mr. Kumar believes that they are managing the computing and graphics segment for profitability rather than for maximum revenue or market share potential.
AMD sees the commercial market as the biggest potential driver of growth in its computing business which is why they’ve made investments into the market segment with their PRO series of APUs. Similarly AMD see the professional graphics market as the key to growth for the graphics business. Last quarter AMD believes that they achieved 24 percent market share in this segment and they expect to continue to grow their market share in this segment. The commercial and professional PC spaces clearly represents the biggest opportunities for AMD’s computing and graphics business.
AMD’s outlook on the discrete graphics business is less rosy. Desktop shipments are down but they’ve seen consistent growth in the notebook discrete business throughout the year. This comeback is unsurprising considering that Nvidia almost knocked AMD out of the notebook discrete market with its Kepler generation of mobile GPUs in 2012 and 2013. The recent launch of Nvidia’s Maxwell chips is no doubt limiting AMD’s high-end graphics ASPs. With all of that said Mr. Kumar considers the performance of AMD’s graphics business to be disappointing.
Mr. Kumar also talked about AMD’s efforts to rein in their operating expenses. Since the beginning of 2012 AMD has reduced that number from about 600 million dollars a quarter to a mere 400 million dollars. Of course their revenues have also declined over that period. He also pointed out that the savings were not coming from reduced research and design spending and that AMD was investing heavily in new ARM, x86, and graphics products. Largely the saving are due to shaving redundancies in the computing and graphics business as well as reducing the number of marketing dollars spent. At this point in time AMD doesn’t see the need to continue investing as many marketing dollars into their computing and graphics business as they did in prior years. Clearly we should not be expecting a major expansion of AMD’s marketing efforts in the near future.
Note: The following is for professional and student level subscribers.
Disclosures: Thomas Ryan and Stone Arch Networking Services, Inc. have no consulting relationships, investment relationships, or hold any investment positions with any of the companies mentioned in this report.
Latest posts by Thomas Ryan (see all)
- AMD Launches its Radeon Instinct Server Accelerators - Jun 22, 2017
- A Look at AMD’s EPYC Line-up - Jun 21, 2017
- AMD Details EPYC’s Power Efficiency - Jun 20, 2017
- Intel’s Skylake-X and Kabylake-X Meet Our Scale - Jun 19, 2017
- MIPS for Self-Driving Cars with the I6500-F CPU - Jun 19, 2017