One of the hottest developments to emerge in the semiconductor industry this year is the recent piece of news that the world’s largest chipmaker, Intel Corporation (NASDAQ:INTC), was looking to acquire one of the leading manufacturers of field-programmable gate arrays, Altera Corporation (NASDAQ:ALTR).
The deal, valued at more than $10 billion, could turn out to be Intel’s largest acquisition ever.
Altera shares have soared about 25% ever since news of the rumored merger hit news feeds, while Intel’s shares are up about 3%. Intel is looking to leverage the Altera acquisition to build out its data center group where it can presumably use Altera’s technology in its Xeon processors.
Meanwhile, the merger is likely to have far-reaching repercussions for the largest programmable logic devices maker, Xilinx, Inc. (NASDAQ:XLNX). XLNX has a 50% share of the PLD market, compared to Altera’s 39%.
Intel could turn to Xilinx if Altera’s asking price becomes hard to swallow. Otherwise, the acquisition is likely to trigger counter-acquisitions by other large server chip makers such as Advanced Micro Devices, Inc. (NASDAQ:AMD) which might set its sights on XLNX.
But by far the most important effect is that the acquisition is likely to open up the rather narrow field-programmable gate array market and benefit Xilinx in the process.
Altera Merger Is an Expensive Proposition for Intel
Altera shares are currently trading at multi-year highs after they skyrocketed following the rumored merger. Altera now sports a market cap of around $12.5 billion, up from just $10.3 billion prior to the rally. The company has an enterprise value of roughly $11 billion.
Assuming Altera’s management wanted at least a 20% premium to its enterprise value for the merger, Intel would have to fork out about $14.2 billion to buy the company. The last time Intel reported its financial details, it had $14.1 billion in cash. The Altera acquisition would drain its entire cash reserves, meaning a good part would have to be financed through debt.
Keep in mind, however, Intel spends more than $10 billion in CapEx each year, which greatly limits how much debt it can take on.
So, why not look for a better alternative? XLNX has an enterprise value of just $9.6 billion — 15% lower than Altera’s reading, yet Xilinx finished fiscal 2014 with a revenue of $2.38 billion — 23% higher than Altera’s revenue.
Intel would pay almost 25% less for Xilinx yet get 23% more on the top line. Even if Intel bought Altera at zero premium to the current value, XLNX would still offer a better value proposition. Barring any other considerations, I think the math is pretty compelling.
Counter Acquisitions
Assuming Intel goes ahead and acquires Altera, I think we are likely to see XLNX being acquired sooner rather than later.
Intel intends to use Altera’s FPGA chips in its Xeon processors. You can bet that the ARM camp, led by Advanced Micro Devices, will not be sitting on its laurels either. ARM chips are used in microservers, which are often touted as the server of the future due to their low power consumption, flexibility, and low cost.
Using FPGAs to accelerate hardware is not exactly a new technology. The technology has so far had limited applications, but that fact is likely to change quickly. Microsoft Corporation (NASDAQ:MSFT) is using FPGAs in its datacenters to make Bing Search a lot faster than Google Inc’s (NASDAQ:GOOG). And Intel managed to produce a hybrid FPGA/Xeon E5 processor last year.
XLNX remains a bit ahead of the curve, having licensed ARM Holdings plc (ADR) (NASDAQ:ARMH) Cortex processor IP back in and 2009.
To give microservers a better chance of competing with Intel’s microserver chips, AMD and other ARM chip makers would have to at least match their specs. Acquiring XLNX would bring the technology in-house and help AMD to compete better with Intel.
Opening up the FPGA Market
The PLD market is currently rather narrow and accounted for just 2% of the $300 billion semiconductor market in 2013. The FPGA market was estimated to be worth about $4.6 billion in 2014, and is projected to grow at 8.1% to reach $8.5 billion by 2020.
Since Xilinx and Altera share a virtual duopoly of this market with close to 90% market share, and both are pure-play FPGA makers, the two companies are likely to see their revenue grow as fast as the FPGA market.
This estimate is conservative, however, because it probably doesn’t take into account the fact that future Xeon processors, and maybe ARM chips as well, are likely to soon start integrating FPGAs in the architectures.
If both the Xeon and ARM camps start integrating FPGAs in their respective architectures on a large scale, my guess is that the FPGA market could easily grow at 10% to 12% through 2020. XLNX will, of course, be the biggest beneficiary of this accelerated growth.
The Bottom Line
It appears that the Intel-Altera merger could turn out to be positive for XLNX in any number of ways. The company can end up being the acquisition target for Intel, or be acquired by an ARM chip maker, or simply benefit from a bigger addressable FPGA market.
Xilinx should end up winning whichever way the die rolls.
As of this writing, Brian Wu did not hold a position in any of the aforementioned securities.
