There were two interesting mergers in play this year with semiconductor companies. The big one was the proposed NVIDIA ARM merger, but also interesting was the AMD Xilinx merger. The NVIDIA ARM merger had no chance of being successful because NVIDIA’s size and ARM’s importance turned the effort into a regulatory approval nightmare. The AMD Xilinx merger, which has passed regulatory approval, will close in the first quarter of 2022.
The difference between the two deals is that both AMD and Xilinx fall below the radar when it comes to mergers, while NVIDIA and ARM are far more powerful, and biotech has a bad name now. This suggests that large companies like NVIDIA will need to get more creative or less ambitious with their acquisitions until the regulatory environment changes, if it ever changes, while smaller firms can continue with business as usual.
Let’s talk about this new dynamic which is beginning to significantly hamper the ability of large dominant companies to create mergers and what might be done to accomplish similar goals more successfully.
AMD vs. NVIDIA
AMD is largely seen as non-threatening because it’s much smaller than NVIDIA and seemingly more focused on PCs and servers. AMD has historically played in Intel’s shadow, but in recent years it has been able to compete with Intel on a more even playing field. Intel is still considered to be the larger and more powerful player, making any acquisition by AMD appear less troubling.
Over the last couple of decades, NVIDIA has become a dominant power in a variety of industries. It is critical to advanced work in supercomputers, AI, autonomous robots and vehicles, and has arguably eclipsed and clearly outperformed Intel, which was once the industry giant, in recent years. For anyone concerned by large company dominance, NVIDIA is one of the scariest companies in the segment, a fact that wasn’t lost on regulators.
AMD’s acquisition of Xilinx appears to make the company more competitive against giants like Intel and NVIDIA while the acquisition of ARM by NVIDIA could have turned NVIDIA into a company that could even give Apple a run for the money. Even though NVIDIA promised to leave ARM alone and the influx of cash from NVIDIA would have likely made ARM stronger, the risk that the result would create a competition killer was simply too great for both ARM licensees and regulators.
It is interesting to note that AMD might have been able to acquire ARM successfully, but AMD’s tight focus makes that acquisition very unlikely. But, going forward, AMD is more likely to be able to accomplish acquisitions than NVIDIA, providing AMD with a situational competitive advantage for the time being.
NVIDIA’s long term
Any acquisition that a large company like NVIDIA makes has a high risk of failing. NVIDIA is hardly alone. Any of the major companies attempting acquisitions right now are likely to find that path closed to them. This means they need to find alternative paths to accomplish the same thing. This could include strategic partnerships, extensive licensing deals, exclusive collaboration efforts, and even building their own division by headhunting from the company they wanted to acquire (this has risks as well but should still be far easier than getting an acquisition approved).
In the NVIDIA/ARM case, ARM is a licensing entity and clearly wants to partner with NVIDIA. That partnership, if structured correctly, could give NVIDIA more than it would have gotten with a highly regulated merger, like freedom from oversight. It could also buy ARM shares once the company IPOs and get on the ARM board for even greater influence, though they’d have to be careful to avoid the appearance of driving anti-competitive decisions.
Finally, NVIDIA could fund an ARM competitor it builds from scratch using its own talent blended with engineers it acquires from CPU companies or ARM itself. For NVIDIA’s long-term goals, this could be even more powerful because it wouldn’t have to deal with ARM licensees, many of whom are also very large, moving to block its efforts.
Right now, being a large successful tech company, particularly in the U.S., is becoming a regulatory oversight nightmare thanks largely to the behavior of companies like Google, Amazon and especially Facebook which arguably fueled this problem through horrid behavior.
The best way for these large companies to acquire critical assets is no longer through buying the related company because that purchase is now more likely to be blocked. Smaller, less threatening companies like AMD can still play here, but large companies will be aggressively blocked.
These larger companies need to think through their merger goals and find more creative ways to accomplish the same task. Ironically, given most mergers fail to meet their objectives, this alternative approach might turn out to be far more successful than a merger because they focus on results, while a merger must first focus on getting the merger done.
So, the path that NVIDIA has been forced to take has a decent chance of being more successful than the merger path it first attempted. And, sometimes, there are advantages to being small and non-threatening or taking the path less traveled.