Nvidia is evaluating increase its portfolio of suppliers with the addition of a rather curious location for develop a part of its chips: the Intel factories. This has been recognized its CEO, Jensen Huangat a virtual press conference held this past Wednesday as part of its GTC 2022 event. Huang, who has also assured that Intel is interested in using its factories, has highlighted that they are «very interested in exploring the possibility».
This step, if taken, would serve Nvidia to increase its production, which would not hurt at all to cover the significant growth in demand for its GPUs by gamers and activities such as cryptocurrency mining. This, together with the shortage of chips, has led to a rise in the prices of GPUs, which has brought important profits to their manufacturers, who are unable to cover all the demand. In the case of Nvidia, which does not have as many problems as other companies in terms of component supplies, the increase in demand and prices have led to record revenues.
Faced with this situation, the company is evaluating the options it has to increase its production, for which it wants to expand its supplier portfolio, in addition to adjusting its products to be able to use a range of components that are manufactured with both older process nodes and with more recent, which will allow you to sell several generations of GPUs at the same time.
In addition, the company is trying to reserve as much third-party factory capacity as it can for the future. In a few years, Intel will have considerably more capacity in this sense, since it is going to build new plants in the United States and Europe, which undoubtedly increases Nvidia’s interest.
That Nvidia can even consider the possibility of manufacturing its components in Intel factories may be shocking to many, since according to Tom’s Hardware Intel is direct competition with Nvidia in both CPUs and GPUs. But it’s not something that worries Huang, who explained that Intel and AMD have known about Nvidia’s secret roadmaps for several years, so he’s not too worried about sharing more information.
Of course, if both companies finally reach an agreement, the production of Nvidia chips in Intel factories would not start overnight. Huang has recalled that deliberations on the use of factories take time, and intentions alone are not enough because you have to “couple technologies, business models have to be aligned, and the same happens with the operations processes and the nature of both companies. It takes quite a bit of time, as well as numerous and in-depth deliberations. We are not buying milk, but integrating our supply chains. Our deals with TSMC and Samsung over the past few years took years to bear fruit. That’s why we’re open to taking Intel into consideration, and I’m delighted with the efforts they’re making.«.
The Register reminds that Huang has a strong working relationship with Intel CEO Pat Gelsinger, who before taking his current role was at the helm of VMware, a company that has become a very strong leg of Nvidia’s data center strategy. Nvidia now sees Intel as a partner, and Huang thinks Gelsinger’s strategy of focusing on manufacturing is the right one. Of course, as we have mentioned, for the agreement to come to fruition, it is necessary to align the technologies of both companies. as well as their business models and capacity.