- Intel has finally closed its huge merger with FPGA maker Altera.
- The merger is expected to have a huge positive impact on Intel's data center growth.
- Altera will also help Intel to remain ahead of the ARM camp and drive Intel stock.
After what appeared like a botched merger attempt, Intel’s record $16.7B buyout of FPGA maker Altera is now officially on the books. The Altera purchase has become Intel’s largest acquisition to-date, and one of the most significant semiconductor M&A activities in 2015. Altera will become part of Intel’s Programmable Solutions Group, or PSG. Dan McNamara, previously Altera’s general manager of Embedded Products division, will become PSG’s head. Intel says PSG will work closely with its Data Center Group that comprises of CPUs and network processors as well as its IoT Group that consists of embedded CPUs and other related products.
Intel (NASDAQ:INTC) expects the Altera deal to be accretive to its cash flow and non-GAAP EPS in 2016 but dilutive to GAAP EPS as well as its M&A-related costs.
Just how accretive will Altera be to Intel’s 2016 earnings? Deutsche Bank’s Ross Seymore estimates that Altera will account for ~9% of Data Center Group Revenue and ~3% of Intel’s overall revenue in 2016. But perhaps the most positive aspect of the merger is that most of Intel’s DCG growth will now come from Altera. The analysts raised their Intel PT by $1 to $37 saying:
"We now estimate Intel’s DCG revs to grow +21% y/y, up from +10% y/y previously as ALTR accounts for ~53% of y/y DCG rev growth. We model 2016 gross margins to benefit slightly (62.4% from prior DBe 62.2%) with opex increasing +2.7% from prior DBe to $20.9b ($20.4b previously). Our resulting 2016 EPS goes to $2.40 from $2.35 previously."
This is a very important aspect of the merger since Intel’s DCG, its most important growth driver, has slowed down lately. During the third quarter of 2015, Intel’s DCG revenue grew by just 12%, or about half of the 25% growth recorded by the division during the previous year’s comparable quarter. Intel blamed a weak macroeconomic environment for the slowdown saying:
Both the data center and 'Internet of Things' businesses will also exhibit strong growth. But the annual growth rate for these businesses will be lower than expectations at the beginning of the year as a result of weaker-than-expected macroeconomic growth. We now expect the data center business to grow in the low-double-digits versus the prior forecast of approximately 15 percent.
Staying Ahead In The Data Center Race
Intel plans to marry its Xeon server processors with Altera’s FPGAs (Field Programmable Gate Arrays) for workload acceleration. Intel says that the first paired solutions will arrive as early as during Q1 2016, and estimates that 30% of data center servers in 2020 will contain FPGAs.
So that’s basically the reason why Intel decided to buy Altera for a huge 40% premium. The company sees a future where FPGAs will play an increasingly bigger role in the data center. Currently, there are only three FPGA manufacturers: IBM partner Xilinx with 50% of the market; Altera with 39%, and Lattice Semiconductor (NASDAQ:LSCC) with the rest. FPGAs account for just 2% of the semiconductor chip market, so there isn’t that much to go around. Intel’s acquisition of Altera is as much defensive as it is offensive.
The ARM camp has been ahead of Intel as far as using FPGAs in data center server processors goes. ARM manufacturers discovered that pairing server CPUs and FPGAs resulted in a pretty dramatic performance gain with just a minor increase in power consumption. Microsoft (NASDAQ:MSFT) and Baidu (NASDAQ:BIDU) are some of the large tech names that have been experimenting with Xilinx (NASDAQ:XLNX) FPGAs to accelerate workloads for their respective search engines. Both companies have reported impressive performance by pairing FPGAs with software instead of using software-only solutions.
ARM had managed to grab about 25% of date center market share about a decade ago by producing low-powered server processors with impressive performance. Intel wants to avoid a repeat of that. Intel has bought Altera to bring the technology in-house and avoid the company being acquired by an ARM competitor.
The Altera acquisition therefore seems to be a highly strategic one. Not only will it have an immediate positive impact on Intel’s floundering Data Center business, but will also help Intel maintain its supremacy over ARM in the data center. Intel recently said that it plans to start relying more heavily on acquisitions instead of developing all its technologies in-house as the case has been in the past. Investors can only hope that Intel’s upcoming acquisitions will be in Altera’s mould, and lift Intel stock.