Xilinx and AMD: an inevitable match?

The Intel-Altera deal may force the hand of the two, but would that be a good thing?

With the Intel-Altera nuptials on, their competitors may find themselves in a shotgun wedding. There is now speculation and insistence that Xilinx, Altera's much larger competitor, should buy AMD, Intel's much smaller competitor.

In the FPGA world, it could be Xilinx is the Intel and Altera is the AMD but the gulf isn't as wide. Xilinx has 60% market share to Altera's 40%, but is viewed as the market leader with a broader array of processors and more customers.

We really don't need to belabor the Intel-AMD differences, do we?

So with Intel and Altera hooking up, people are now looking at Xilinx and thinking it needs to do something similar and there's really only one x86 option. A Seeking Alpha writer was the first to float the notion of a merger between the two, making some valid points.

Once Intel is in possession of the Altera product line, it will be able to cheaply produce the chip and drop the price, drastically undercutting Xilinx, and not hurting itself in the process. However, the author failed to note that Xilinx won't get this kind of advantage with AMD, since AMD is fabless.

Since the action appears to be in the server market, AMD would open up the opportunity for Xilinx to have an x86 and FPGA option. Again, the author missed a rather important detail: AMD's server position is miniscule; in the single digit range, according to Mercury Research. While x86 has 73% of the server market, Intel owns virtually all of it.

Somewhat overlooked is AMD's GPU position. AMD may be losing ground to Nvidia but Nvidia is to GPUs what Intel is to CPUs, and AMD's struggles in recent years only allow a ferociously competitive company like Nvidia to open up a lead. So AMD still have a GPU story, for now. I am concerned that the GPU in Intel's forthcoming Skylake CPUs may close whatever lead AMD has over Intel right now.

Barron's also raised the issue in discussing an analyst who feels AMD is a takeover target and Xilinx is a suitable buyer. Gus Richard, formerly with Piper Jaffray and now with Northland Capital Markets, has initiated coverage of AMD with an Outperform rating and a $5 price target, which is awfully generous for a stock hovering at $2.29 as of this writing.

"Investors have decided that AMD has hit the iceberg and they are waiting for the ship to sink," he said, according to Barron's. "However, we believe that AMD still has time to avoid the collision, and if all else fails investors will be bailed out by the lifeboat of M&A."

Both Seeking Alpha and Richard took more of an investor's look than a technological look, although both did mention Xilinx could use AMD as a server/data center play. Again, that overlooks AMD's pitiful position in the data center, but even that could change.

Next year, AMD will begin rolling out a new microarchitecture, called Zen, which on paper looks like it will significantly close the gap with Intel. I have no idea if it will be enough to get server vendors buying again because it's way too early. From the early leaks it looks good and seems to be generating some excitement among tech nerds.

Richard also floated the notion of a Chinese buyer for AMD, but we've been through this before. That is just a no-go and he should know that. Intel holds a license agreement with AMD on x86 patents and no way, no how, is x86 IP going to China. Even if Intel were stupid enough to agree, the DoJ certainly won't, especially after this most recent security breach instigated by the Chinese.

A lot of people seem rather anxious for AMD to be bought. I agree with the notion of some analysts that new CEO Lisa Su does not want to just shine the place up for a sale, she wants to make the company competitive. I wish her luck in that regard, but if Xilinx makes an decent offer, by all means, take it.

ITWorld DealPost: The best in tech deals and discounts.
Shop Tech Products at Amazon