AMD (AMD, Financial) has been the hottest technology stock this year with a 177% gain year to date. The company had been in the red for the past couple of years given the challenges with the previous architecture, Bulldozer, and the fact that it was using a lagging node in contrast to Intel’s(INTC, Financial) process technology. However, the recent uptick in the stock price came from the company’s semi-custom ventures, IP monetization strategy through deals with joint ventures with THATIC and ATMP, launch of Polaris based GPU and the positive sentiment regarding the upcoming Zen Chips. Further, as console manufactures are getting prepared for the next iterations of consoles, things are starting to look rosy for AMD. Let’s analyze the optimistic sentiment about AMD and whether there exists something to the contrary.
The Bull View
Quarterly results are getting better for AMD. The company posted revenue of $1.03 billion during the second quarter of 2016, an increase of 9% on a year-over-year basis. Revenue for third quarter is expected to grow by 18% on a year-over-year basis. The increase in revenue was the result of higher notebook processor sales, GPU sales and a higher sale of custom SoCs. Upcoming quarter growth will be supported by the revenues from Polaris GPU. Note that AMD is the first vendor to address the mainstream market with 14nm Polaris while NVIDIA (NVDA, Financial) is planning to launch 14nm parts by the end of 2016. According to AnandTech,
“Looking at the overall performance picture, averaged across all of our games, the RX 480 lands a couple of percent ahead of NVIDIA’s popular GTX 970 and similarly ahead of AMD’s own Radeon R9 390, which is consistent with our performance expectations based on AMD’s earlier hints. RX 480 can't touch GTX 1070, which is some 50% faster, but then it's 67% more expensive as well.”
They go on to conclude that, right now, AMD has the mainstream market for the taking. Overall, the small window of process technology lead combined with the fact that AMD is pricing its products aggressively will bring material gains in GPU revenues during the current quarter.
Secondly, in embedded systems, AMD is better positioned than Intel and NVIDIA. The company has both the CPU and GPU capabilities, which can bring cost synergies and result in competitive pricing for custom SoCs and embedded systems. That is why we see AMD is now the go-to supplier for console manufactures. Going forward, the shift to 14nm SoCs will put AMD in an even better position. One other thing is the domino effect due to the launch of Zen. All AMD processors will be coupled with either an integrated graphics or discrete cards. If Zen steals market share from Intel in the Ultra-book market, this will also boost the GPU revenues of the company.
Note that Zen is expected to be head-to-head with Intel’s offering, as it will be on the same process node for some time. Zen is expected to result in a 40% IPC increase. Zen can result in server market gains if AMD prices it aggressively. According to AnandTech, the prospect of Zen playing an active spot in enterprise seems very real. Lisa Su mentioned that the company has secured several key design wins with global OEMs for Zen-based server CPUs.
Power efficiency will be the key competitive advantage for AMD. The company was lagging behind in node technology and had to innovate in order to cope in terms of power efficiency. They had to find new ways to increase efficiency, as moving to the newer node was not at AMD’s disposal. On the other hand, Intel did not have to worry about efficiency, as they were always on an advanced node, which took care of efficiency gains. The point is that AMD is able to carry forward the innovative methods they developed, for efficiency gains. This leads me to believe that AMD will be able to outpace Intel in terms of efficiency on a similar node. Therefore, they have a distinct competitive advantage as far as Zen is concerned. 
AMD investors’ presentation
Another positive for the company includes the IP monetization strategy. Joint venture with THATEC exposes the company to the Chinese server market, which is the world’s second largest server market, $293 million in cash and income generation is expected from the joint venture. The joint venture with NFME resulted in net cash proceeds of $320 million, but what is interesting is that this will result in significant CAPEX reduction. Licensing of IP seems to be a good strategy, think Qualcomm (QCOM, Financial). But, AMD have the x86 agreement with Intel that does not allow licensing the technology to third parties. The terms of the venture are not very clear. Therefore, a legal battle with Intel might be in the cards.
Further positives include cost discipline and clarity about the capital structure. OPEX declined 26% since the second quarter of 2013. And, debt repayments are not due until 2019. Therefore, the bankruptcy arguments no longer holds true. The company is in the clear for now.
Investors’ presentation – May, 2016
The Bear View
But, all these positives do not mean that AMD is going to dominate Intel going forward. There are several concerns that force me to believe that AMD will face difficulties over the long run. Note that I have been an AMD bull since 2014 and all the positives have the ability to pan out. But, being a bull or a bear very much relates to the price for a given stock. At $2-$3, it made sense to go long on AMD given all the positives mentioned above. But, the stock is touching $8 now and it is not a good idea to hop on the momentum bandwagon, if one is investing from a long term perspective.
14nm is Expensive
The average IC design cost for a 14nm chip is about $80 million, compared to $30 million for a 28nm planar device, according to Gartner . AMD usually competes on price. And, this has been the theme for upcoming Zen in the server markets that AMD will steal market shares amid aggressive pricing of its server CPUs. Cutting prices will pressurize margins as AMD has to pay for 14nm manufacturing. To compete on price, AMD has to forgo margin. Competing on price on the 28nm node was plausible as 28nm was much cheaper than the advanced nodes. At 14nm, pricing will certainly be a challenge for AMD.
10nm – A Headache for AMD
Intel’s 10nm is a problem for AMD. Once 10nm based chips are out, AMD will be in the same position, that is to compete in the mid and low performance range of the market. It is not very likely that AMD will move to 10nm at all. The company will most probably skip it because it has to take cost driven decisions amid the limited resources it has to play with. Previously, AMD jumped from 28nm to 14nm, skipping the transition node to 20nm.
Further, there are rumors that Global Foundry is skipping 10nm to focus on 7nm. Thereby, it is more likely that AMD will skip 10nm. The implications can be that AMD will lag behind Intel in terms of process node and has to settle for mid-to-low end of the market. Note that 10nm will ship by year’s end, roughly a year ahead of 7nm. Especially in sever and data center space, advanced node is a key advantage as lagging node technology can add to the total cost of ownership.
Execution is Yet to be Seen
All the claims about Zen are on the paper right now. We have seen the execution problem happening in the past with the Athlon Chips, which were also designed by Jim Keller. They [AMD] really have to execute,” said Martin Reynolds, an analyst at Gartner Reynolds in an interview.
“Obviously, we’ll have to wait until we see the final product. All of the changes they made in the architecture are the things they need to fix. We don’t have enough information to know if it will beat Intel.”
Especially in marketing, AMD does not have the financial resources to compete with Intel in marketing terms. Being good at marketing is a key to sell. Most of the mainstream users, i.e. the public, do not know much about the technical details of the product and the company with marketing firepower gets the most of the market share. This is the area AMD has to focus on once it launches the Zen in early 2017.
Final Thoughts
To sum it up, AMD has a window of opportunity until the second half of 2017 as Intel is going to roll-out 10nm parts by then. However, even a marginal gain in the server space through Zen can bring material revenues to AMD. Over the long-run, the node transition story seems to remain the same for AMD/Intel i.e. AMD using a lagging node and competing on prices. And, this makes it difficult for AMD to compete against Intel in the high-end market. Therefore, it is not a good idea to bet on AMD in the long-run, given the current stock price. In the short-run, the stock can go either way, based on how Zen performs compared to the parts from Intel and whether AMD’s marketing is able to attract new customers.
From an investment perspective, shorting the stock is not a good idea given Polaris related market share gains, consoles refresh cycles and the promise of Zen. Investors should lock-in the profits through puts. Outright selling is also not a very good strategy given positive news in the short-term can send the stock north of $10. However, over the long run, $10 seems questionable for AMD.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
Start a free 7-day trial of Premium Membership to GuruFocus.
