Intel (NASDAQ:INTC) could be "pondering the acquisition" of Advanced Micro Devices (NASDAQ:AMD) according to a recent EE Times report, citing "an intriguing rumor" circulating at CES. In theory, a takeover would cement Intel's position as the world's top maker of x86 CPUs, expand its portfolio with AMD's GPUs and embedded SoCs, and possibly conclude its CEO search by appointing AMD CEO Lisa Su as its new leader. But in reality, an Intel takeover of AMD could never happen for five simple reasons.
1. Regulators would never approve it
Intel controls about three-quarters of the PC CPU market and more than 90% of the data center CPU market. Its only viable competitor in both markets is AMD, the only other major producer of x86 chips, so the takeover would transform the duopoly into a monopoly.
It's highly unlikely that antitrust regulators in any country would approve the merger, since it would give Intel absolute pricing power and control over the consumer and data center CPU markets.
2. The price tag
Intel's $17 billion purchase of Altera and its $15 billion takeover of Mobileye are its two largest acquisitions to date. AMD has an enterprise value of about $20 billion. Factoring in a 50% acquisition premium, it could cost Intel up to $30 billion to seal the deal.
That would be a massive purchase for a company that finished last quarter with just $3.4 billion in cash and equivalents. Intel is also still gradually paying off the $25 billion in debt it accumulated from its previous deals.
Intel plans to finish fiscal 2018 with a non-GAAP free cash flow of $15.5 billion, but it already deployed most of that cash ($12.6 billion in the first nine months) on buybacks and dividends. Therefore, if Intel is still looking for acquisitions, potential targets should be much smaller than AMD.
3. Partnerships with AMD make more sense
If Intel buys AMD, it would inherit its discrete GPU business, which is struggling with competition from NVIDIA (NASDAQ:NVDA) and the loss of cryptocurrency mining revenues.
Those headwinds are cyclical, but Intel already partnered with AMD last year to launch Kaby Lake G, a laptop chipset that merges an eighth-generation quad-core Intel CPU with AMD's custom Radeon RX Vega M GPU.
This supported rumors that Intel could license AMD's graphics technology to replace its expired cross-licensing deal with NVIDIA, but Intel denied those claims. Nonetheless, partnering with AMD for GPU-integrated chipsets like Kaby Lake G makes more sense than buying AMD outright.
4. It doesn't make strategic sense
Intel's acquisitions of Altera and Mobileye expanded its reach beyond the "mature" markets of PC and data center CPUs. Altera provided it with programmable chips (FPGAs), which can be customized for a wide range of industries. Mobileye gave Intel a leading position in the ADAS (advanced driver-assistance system) market and computer vision chips for autonomous cars.
Intel also acquired smaller companies, like the computer vision chipmaker Movidius, the machine learning firm Nervana, and the IoT (Internet of Things) and ADAS company Yogitech. These moves all indicate that Intel wants to evolve by expanding its reach into newer higher-growth markets, but buying AMD would merely increase its exposure to the older PC and data center CPU markets.
5. There are better takeover targets
Lastly, the recent sell-off in semiconductor stocks makes many smaller chipmakers look like better takeover targets than AMD. Two potential targets are Ambarella (NASDAQ:AMBA) and Cypress Semiconductor (NASDAQ:CY).
Ambarella, which makes image processing SoCs and computer vision chips, has an enterprise value of about $860 million. Acquiring it would take out one of Mobileye's chief competitors, and expand its reach in cars, drones, and other connected camera markets.
Buying Cypress, which makes niche memory chips and embedded chips for cars, industrial machines, consumer electronics, and IoT devices, would expand Intel's portfolio beyond x86 CPUs and complement its non-volatile memory chip business. Cypress has an enterprise value of just $5.5 billion.
In other words, it would be smarter for Intel to acquire these smaller players than to make a massive bid for AMD, which would inevitably trigger an antitrust probe, increase its exposure to older chip markets, and boost its long-term debt. Therefore, investors should ignore this latest rumor and focus on Intel's other challenges -- like finding a new CEO and overcoming its chip shortages.