Microsoft (MSFT -0.62%) is rumored to be interested in acquiring chipmaker AMD (AMD -0.74%), according to recent reports from Reuters and KitGuru. The takeover talks reportedly started a few months ago, although additional details are currently unknown.
AMD currently has a market cap of $1.8 billion, and Microsoft has $95.3 billion in cash, so it would be an easy purchase for the tech giant. But before investors get too excited about this buzz reviving AMD stock -- which has fallen more than 40% over the past 12 months -- they should understand that these rumors are unlikely to bear fruit.
Possible reasons for buying AMD
AMD has been losing ground to Intel (INTC 0.10%) in x86 processors and Nvidia (NVDA -0.35%) in graphics cards for years, yet it recently found a comfortable niche with SoCs (system on chips) for gaming consoles. AMD currently provides the SoCs for Sony's (SONY -0.36%) PS4 and Microsoft's Xbox One -- the two top consoles of the current generation.
Microsoft and Sony both pay AMD roughly $100 per console for those SoCs, which merge the CPU and GPU on a single chip. Microsoft has sold 12.6 million Xbox Ones since late 2013, which equals $1.26 billion in payments to AMD. Therefore, buying AMD to save the gross profits on those payments (estimated to be about 20%) sounds like a good idea. Sony would be ironically forced to pay Microsoft to keep using AMD SoCs if Microsoft bought AMD, and Microsoft could further lower the price of the Xbox One to undercut the PS4.
AMD claims that its graphically powerful SoCs are well-suited for augmented reality (AR) and virtual reality (VR) applications. If Microsoft acquires AMD, it could install those SoCs in the HoloLens AR headset and similar devices. AMD also owns a software suite for VR headsets, known as LiquidVR, which reduces the lag in VR image processing that can cause motion sickness. That software, which only works with AMD hardware, could be useful if Microsoft expands its presence into the VR market.
But here's the problem
Microsoft and AMD certainly have some aligned interests, but AMD has huge flaws that could offset any synergistic gains. AMD's business is split into two parts -- the Computing & Graphics (x86 chips, GPUs, and graphics cards) division and the Enterprise, Embedded, and Semi-Custom (SoCs, embedded chips) one.
Last quarter, revenue at Computing & Graphics, which accounted for 52% of AMD's top line, fell 38% annually to $532 million, and its operating loss widened from $39 million to $107 million. This was due to lower PC sales and competitive pressure, which forced AMD to lower prices to preserve market share.
AMD's EESC business, which generated the remaining 48% of its top line last quarter, fared slightly better. Sales slipped just 7% annually to $498 million, but operating income dropped 47% to $45 million. AMD blamed "seasonally lower sales" of semi-custom SoCs, hinting that rising sales of consoles wasn't enough to lift the entire business. Therefore, it's impractical for Microsoft to inherit all those problems just to cut the cost of the Xbox One or gain an early edge in the fledgling AR/VR markets.
A partial sale wouldn't work
To cut off the burnt crust, Microsoft could theoretically buy AMD's EESC division and leave the Computing & Graphics unit with AMD. Unfortunately, AMD probably wouldn't agree to that partial sale for two reasons.
First, the SoCs which Microsoft needs are based on technologies developed for its server microprocessors and high-end graphics cards. Selling one unit without the other would cause a major fracture in R&D. Second, the EESC business is AMD's only way to offset the ongoing losses at its Computing & Graphics division.
But even if AMD sold the EESC business to Microsoft, the division's lackluster bottom-line growth would weigh down Microsoft's earnings, which are already under pressure from offering free/discounted Windows licenses and year-long trial offers of Office 365. Analysts are already expecting Microsoft's earnings to drop 8% this year, and buying AMD -- either as a whole or in bits -- would likely lead to downward revisions.
Shop elsewhere, Microsoft
AMD's market cap is basically pocket change for Microsoft, but inheriting all of AMD's problems simply wouldn't be worth the effort. Over the past two years, Microsoft has bought plenty of smaller companies to boost its cloud business, which had an annual run rate of $6.3 billion last quarter. Therefore, it makes more sense to buy smaller CRM, machine learning, or analytics firms to grow its cloud business than it does to acquire a fallen chipmaker like AMD.