Investors continue to think about AMD (NYSE:AMD), Intel (NASDAQ:INTC) and Nvidia (NASDAQ:NVDA) in terms of historical rivalries. As shown by Q4 2013 results and plans for 2014, these three companies are increasingly pursuing divergent market sectors where there is less mutual competition. This has already resulted in higher revenue for each company, despite the flattening of PC demand, a trend that should continue in 2014 and beyond.
The biggest of the companies, Intel, has struggled the most to redefine itself in an era of stagnating traditional desktop and notebook PC unit shipments. Intel has the money to continue to pursue the smartphone and tablet markets, but other obvious opportunities are in server chips for cloud computing and licensing its advanced process technology.
Nvidia just reported a great December quarter and made its segment-specific plans more clear. Automotive semiconductors will be a big story for Nvidia going forward, at first for infotainment and later for smart (or at least smarter) cars. PC gaming GPUs will remain a core strength for a while, but Android gaming and server GPUs will play an increasingly important role in generating revenue and profits.
AMD has already switched its future emphasis from the traditional PC market to gaming consoles, embedded SoCs and niche server chips. AMD continues to go head-to-head with Nvidia in PC GPUs and with Intel in PC CPUs (or increasingly, SoCs combining the CPU and GPU), but these are large markets where AMD still has enough market share to justify the competition.
Intel and Nvidia are both cash-rich, and return that cash to shareholders through buybacks and dividends. AMD is cash-poor, and will likely require a couple of years of good execution on its new strategy before being able to return cash to shareholders. Of course, in the rapidly changing world of technology, there are perils for all three companies, including competition in some niches and the usual possibility of failing to execute.
Before going a bit more in-depth into the details of my divergence hypothesis, take a look at revenue and profit for each company, Q4 2013 vs. Q4 2012 [note Nvidia is on a fiscal year ending January 26; AMD and Intel on calendar years], using GAAP net income:
|y/y % change||Q4 2013 |
|Q4 2012 |
AMD looks like the growth leader, but 2014 growth will be slower as chips for new console models were the main drivers in 2013. Intel's revenue growth was anemic, but net income growth was better, reflecting continued success in dominating the server chip market. Nvidia looks less impressive, but within those numbers, there is a shift to new markets that make Nvidia much less dependent on the high-end gaming GPU market.
While areas of competition between the PC semiconductor trinity remain, they are considerably lessened from the previous decade. Intel and Nvidia are not competing in gaming consoles, with minor exceptions. AMD has stopped competing in the high-performance CPU category. In mobile, Nvidia chose to enter the ARM+GPU market, not the x86+GPU mobile market, where so far both AMD and Intel have been playing weak hands.
Each of these companies brings advantages to its new strategy. In some cases, there are new sets of competitors. Nvidia and Intel are competing against Qualcomm (NASDAQ:QCOM), Apple (NASDAQ:AAPL) and many others in the smartphone and tablet markets. It is tough competition, market share gains are still a possibility even as the industry consolidates. In the consolidation phase, margins tend to shrink (and then expand when the dust has settled), so having plenty of cash can make the difference between survival and extinction.
Many companies would like to break into the server CPU market with designs based on the new 64-bit ARM CPU architecture. AMD has an advantage from having a long history of making x86 server chips, and a disadvantage from coming late to the ARM architecture. No one is really sure how well the ARM-based servers will compete against x86 servers. Nvidia, while not traditionally a CPU company, has a great deal of experience selling GPU cards into the server market. It will be interesting to see if Nvidia sticks to GPUs for servers or makes a play for the CPU side as well.
There is a price that has to be paid for changing strategies: R&D and other operating costs, plus capital expenditures. Since Nvidia and AMD are both fabless, their capital expenditures are much less relative to Intel, which appears to have invested in a surplus of production capacity. AMD lost money during its transition to CPU+GPU chips, including those used for consoles, putting it at a disadvantage. Nvidia was able to maintain a profit stream while transitioning to its new Tegra and cloud strategies, even returning money to shareholders during the process.
These strategic changes will take some time to play out and are against a backdrop of rapid technological change, notably the shift from local-PC based computing to the cloud+client model. But there are other factors at work as well: the transition to high-definition video, high-definition screens and generally higher expectations from end-users continues to drive a need for faster graphics and general computing ability, more storage in the cloud and on client devices, and lower electric power consumption.
AMD investors have been waiting for a turnaround for so long that there is naturally much skepticism in the investment community about the new paradigm. The main reason to hope that AMD will succeed is that it is no longer just the underdog in two struggles, against Intel in PC CPUs and against Nvidia in graphics chips. AMD is clearly number one in gaming consoles, a sufficiently large market to serve as a secure base for expanding to adjacent markets. AMD needs to expand into adjacent areas where it is not competing directly with Intel and Nvidia, rather than taking any cash generated from consoles and using it to renew the CPU and GPU wars. AMD can't withdraw from the CPU market, but it can target PC niches where it has an advantage over Intel because of its better integration of graphics processing.
I have been watching Nvidia critically during the last two years [see, for instance, Nvidia's Competitive Outlook], wondering whether brave talk of conquests outside of the traditional GPU market would compensate for the shrinking of the PC-discrete GPU market. Results for the quarter ending January 26, 2014 tipped me to a favorable view of Nvidia. Revenue in Q4 was up from year-earlier, and news outside of discrete GPUs for PCs was good. Tegra, a mobile chip family based on ARM CPU and Nvidia GPU architecture, has not exactly conquered the mobile world, but it can be found in some smartphones and tablets and has some wins in automotive entertainment systems that could lead to more substantial revenue growth in 2014. So far, Nvidia is beating AMD in selling GPUs into the datacenter/cloud computing markets. These GPUs for the cloud, including the GRID initiative, are a complement to the Intel CPUs that currently dominate the server market.
Intel should be left with the bulk of the x86 CPU market in both PCs and servers. Although non-mobile PCs are not a fast-growth area, they are also not going away anytime soon. Whether Intel can pick up market share in smartphones and tablets remains an open question. Intel clearly made a fundamental strategic mistake abandoning its early experiments with ARM in favor of pursuing x86 for mobile. Intel thought more powerful mobile devices would require x86, but instead, ARM architecture evolved to meet the challenge. Yet, Intel is in a formidable position. It has the most advanced fabrication capabilities and actual capacity in the world. It has cash and securities of $17.9 billion on its balance sheet. At Friday's closing price of $24.75, INTC's dividend is 3.64%.
More than many tech investors, I am generally skeptical about yesterday's winners being tomorrow's winners. I lived through the mass extinctions of mainstream computer companies, mini-computer companies, early (pre x86) microcomputer companies and Internet bubble companies. While there are no guarantees, I now believe that Intel, Nvidia and AMD are all going to survive at least the next leg of the technology race. Nvidia seems to be in the best position to actually thrive. AMD had a great 2013, but has to run the leanest and is still subject to competitive pressures from Nvidia and Intel. Intel is a cash cow, will be facing less direct competition from AMD, and has the resources to explore any new money-making opportunity that arises.
I believe this divergence will continue, and that the result of less direct competition will be better profitability over time. From my point of view, INTC, NVDA and AMD have all become Buys.