I am an NVIDIA employee in a niche of the HPC business. This is not an NVIDIA official opinion.
HPC is nice, but when you hear Jensen getting really excited, it’s not about dominating some niche like that, it’s about a vision of the shiny sci-fi high tech future, and actually delivering the tech to make it real.
So don’t just look at HPC to understand the NVIDIA ambition. Start at edge computing; imagine a world with ubiquitous autonomous robots (cars and drones and otherwise). Think of the onboard chips driving their vision and speech recognition models: That’s a great place for ARM and NVIDIA chips together, whether as one company or two. Watch a recent keynote and see how all the rest of the tech fits into place as part of that: 5G signal processing chips, for instance, something you might gloss over if you’re not in telecom. You don’t need a roadmap to see how it is all connected in support of this world of the future.
(I certainly don’t have the roadmap, either, I just watch the keynotes and help shuffle bits.)
It seems pretty clear that this what they're thinking of. They want to be able to license an integrated architecture that includes power-efficient computing and a powerful ML engine. They've been so heavily investing in this space for a reason.
What I can't figure out is why this is such a big deal to regulators. Nvidia doesn't manufacturer these things (aside from Jetson I believe? Not 100% clear on this). They license IP. And this is IP that I think the world would really like to have.
Currently the only player in this space is Apple. They've built their own integrated silicon with their perpetual ARM license that is now giving them a huge market advantage, and will continue to do so until there is another competitor. The R&D required to compete with a cash-liquid >2.5 Trillion dollar company is just not feasible for any of the other major players at present. Nvidia/ARM opens doors for tons of other companies.
I also think it's foolish to think that Apple won't try to expand this tech offering well beyond personal computers and tablets. They will expand to IoT/Edge devices and services. But the difference is they won't be licensing their IP to other manufacturers, they will be building it themselves (or contracting Foxconn to) and keeping everything in their walled garden.
Guess I'm just frustrated that of all ridiculous acquisitions and anticompetitive nonsense I've seen in the past decade, THIS is the one getting smothered.
> What I can't figure out is why this is such a big deal to regulators.
Because when you own all the IP, you can cut your competitors off by revoking licenses to them, and it'll instantly kill a huge ecosystem from Raspberry/OrangePi to Ampere A1 and everything in between.
I'm not sure nVidia would make such a drastic move, but I'm sure that they'll move strategically to ensure their leadership, which is understandable from a corporate PoV, but it'll be very bad for everybody else.
This is not a big deal, it's a huge deal, and I'm happy that we're here as of today.
nVidia can of course license ARM to embed and/or further improve upon this, or they can use any other ISA or come up with their own. I'm sure they're capable of this, and it'll be much better in the long run for everyone.
> I also think it's foolish to think that Apple won't try to expand this tech offering well beyond personal computers and tablets. They will expand to IoT/Edge devices and services. But the difference is they won't be licensing their IP to other manufacturers, they will be building it themselves (or contracting Foxconn to) and keeping everything in their walled garden.
nVidia's walled garden is not different in any scale when compared to Apple. Considering how friendly nVidia was towards OpenCL, I'm guessing that they'll be at roughly the same distance towards Vulkan for GPGPU applications, keeping CUDA the only possible thing to run with any meaningful performance on their hardware. On the open driver front, they're equally friendly. So it's more like the pot is calling the kettle black here.
HPC is nice, but when you hear Jensen getting really excited, it’s not about dominating some niche like that, it’s about a vision of the shiny sci-fi high tech future, and actually delivering the tech to make it real.
So don’t just look at HPC to understand the NVIDIA ambition. Start at edge computing; imagine a world with ubiquitous autonomous robots (cars and drones and otherwise). Think of the onboard chips driving their vision and speech recognition models: That’s a great place for ARM and NVIDIA chips together, whether as one company or two. Watch a recent keynote and see how all the rest of the tech fits into place as part of that: 5G signal processing chips, for instance, something you might gloss over if you’re not in telecom. You don’t need a roadmap to see how it is all connected in support of this world of the future.
(I certainly don’t have the roadmap, either, I just watch the keynotes and help shuffle bits.)