Nvidia and Intel announced a partnership/investment arrangement that will provide advantages for both players. Indeed, while Intel is more in need of a partnership than Nvidia at this point, that doesn’t negate the fact that Nvidia is getting some real advantages with this partnership and investment as well.
Nvidia announced that it will be investing $5B in Intel through a stock purchase arrangement. For Nvidia, that is a fairly modest investment given its current financial position. For Intel, it’s a major funding opportunity given its troubled financials as of late. But what are both getting from this partnership?
Intel, of course gets a needed boost in financials to invest in its recovery and foundry rehabilitation and foundry service ambitions. For this, it’s committing to work with Nvidia on future data center and PC x86 based chips, and importantly for Nvidia, is adopting the Nvidia proprietary NVLINK protocols. Not generally mentioned in some analysis, this means that Intel has essentially determined its ability to compete head to head with Nvidia in the current large scale AI marketplace, despite its best efforts, have mostly failed. Intel still does have some capabilities to advance its own products in the emerging AI Inference marketplace and edge computing, although Nvidia is also pursuing that space, Nevertheless, it remains a major long term opportunity for Intel. But it also means that Nvidia now has a partner that can provide it with product capabilities and market share in edge, PC workstation and small server environments it may have struggled to obtain otherwise.
For Nvidia, it gets the aforementioned adoption of the NVLINK interconnect, which is a critical technology propelling their higher end AI products. Nvidia already uses a number of Intel Xeon data center chips to power their largest systems, and the x86 chips provide most of the controls and pre-processing that their large scale GPU racks require. Nvidia now gets to obtain accelerated capabilities from those Intel CPUs that will enhance Nvidia performance by potentially getting customized CPUs optimized to the needs of Nvidia. It also means that while Nvidia had toyed with building out its own CPUs, it now seems to admit that working with an established player like Intel is a better long term bet.
While not initially stated as part of this arrangement, a close working relationship between Nvidia and Intel could eventually mean Nvidia utilizing Intel manufacturing for its chips and packaging capabilities, assuming that Intel can indeed get its top of the line processes to work as they plan (14A). Any relationship with Nvidia at this point, while not explicitly talking about the foundry services, should be seen as a possible extension of the partnership in the future, and if it happens, would be a major win for Intel foundry services.
It should be stated that while this arrangement is highly beneficial to both parties there are some losers. In particular AMD may be disadvantaged as a result of this partnership, although they are doing well with their own AI efforts in GPU and CPU. Still, having two major competitors combining their efforts is not exactly a positive outcome for AMD. And with a slew of startup contenders for AI market share, the collaboration of these two behemoths should be seen as a bigger hurdle to jump through for any real market share.
Bottom line: this partnership should be seen as an advantageous arrangement to both Intel and Nvidia, and not just seen, as some might say, a bail-out for Intel. Both get real value and market advantages from the relationship. And ultimately this will make the AI marketplace more expansive and more efficient.
Jack Gold is founder and principal analyst at J.Gold Associates, LLC. With more than 45 years of experience in the computer and electronics industries, and as an industry analyst for more than 25 years, he covers the many aspects of business and consumer computing and emerging technologies. Follow him on Twitter @jckgld or LinkedIn at https://www.linkedin.com/in/jckgld.