Talking about a Nvidia CPU is not something especially new, in fact, for several years now there has been speculation about the possibility that the graphics card giant made the leap into the processor market. Specifically, has been talking about it since 2011, when it obtained a license to use ARM technology with which to design your own integrated based on this architecture. Yes, ARM, it is no coincidence that now Nvidia wants to acquire ARM Holdings to advance much further in that regard.
And it is not, or at least it seems to me, casual the moment chosen to take this step. And it is that, until now, we have always thought that the ARM architecture does not offer the necessary performance to “motorize” desktop systems, whether portable or desktop. However, while waiting to see its first practical applications, Apple intends to demonstrate, with Apple Silicon, that this is not the case, and that an ARM processor may be capable of it. In this context, a ARM-based Nvidia CPU would make all the sense in the world.
And today we know, from DSOG, that just yesterday, at a conference, Nvidia CEO Jensen Huang stated that we may see an Nvidia CPU in the future. It all started when a journalist asked him about Nvidia’s plans for ARM, if your intention was, starting from Neoverse (a family of processors developed by ARM), create a specific CPU development that it would bear the Nvidia signature and that it would aimed at the data center market.
Faced with such a specific question, Huang chose to give a surprisingly broad answer, and that although it pointed primarily to the data center sector, it left the door open to any other type of development. Its purpose is, in collaboration with the entire network of partners and collaborators, to create the broadest possible ecosystem, and to offer various ways of integrating ARM technology in new integrated developments: from the purchase of finished chips to the license of certain aspects of its technology so that other companies can create their own integrated.
Although it is undeniable that the popularity of Nvidia, for many years, has been sustained in its graphics cards and its GPUs, both its hardware and its technologies have had, for many years, a great fit in various fields of the professional sector, such as cryptocurrency mining or artificial intelligence, processes in which the great capacity for calculating floating point operations offered by graphics chips puts them far ahead of CPUs.
However, and although it follows that the first Nvidia CPU would undoubtedly be aimed at the data center and server sector, that you have chosen precisely this moment, in which ARM is about to reach consumer computers hand in hand with Apple, it makes me have serious doubts about whether, finally, the company will not decide to emulate those of the apple and propose ARM-based chips for the big market, whether for computers or consoles.
We may have to wait months, or maybe even two or three years, and of course, it all depends on the approval of the purchase by the regulators. However, and with the proper patience, little or nothing would surprise me that an Nvidia CPU for PC, and that also integrates like a glove with the GPUs of the manufacturer who are in charge of the graphic section of most of them, it is only a matter of time.