Jen-Hsun Huang
Analyst · Matt Ramsay with Canaccord Genuity. Your line is open. Please proceed with your question
First of all, I think I've said this before and I continue to believe it and more so every day, that the datacenter and accelerated cloud computing will likely be the largest opportunity that the Company is following today. Our ASPs there are many, many, many times higher than our ASPs in average. The gross margins are obviously a lot higher, primarily because it's a software business. Now, at the center of that platform strategy is recognizing that datacenters, Internet service providers, cloud service providers, even countries who have their own supercomputing interests, have rather different opinions about what kind of CPUs they use. They don't all want exactly the same instruction set architecture. What we've done with Tesla is, and what we've done with our GPUs, GRID and Tesla, is established a bit of an ISA-neutral attitude about what datacenters can and cannot use, and we believe that people want to use whatever instruction set architecture they would like to use and we've created the ability to support x86, we support Power with our announcements with IBM, a very deep and quite a broad strategic alliance to bring accelerated computing to the Power platform, and also ARM. Of course we care about ARM. ARM is one of the most pervasive CPU ISAs in the world, it is very accessible, it's available to anybody who would like to build CPUs around them. And so, our first philosophy is focus on building a platform for accelerated cloud computing, accelerated computing. Our second is to be ISA-neutral, so that Tesla could be a platform that any service provider or any datacenter can build around. And then third, if it makes sense for us to build CPUs for those platforms ourselves, we'll consider doing so, but it kind of goes in that particular order. Is that helpful?