Sundar Pichai

So there's just a lot of innovation which we are pushing on the frontier, obviously to support our core products like Search, YouTube, Gemini, et cetera. But 50% of the compute goes to its Google Cloud.

So there's just a lot of innovation which we are pushing on the frontier, obviously to support our core products like Search, YouTube, Gemini, et cetera. But 50% of the compute goes to its Google Cloud.

Look, first of all, at a high level, NVIDIA is a phenomenal company. Jensen is awesome. We have been working with NVIDIA now for a very, very long time, and we continue to do so. And we serve a lot of the Gemini traffic on GPUs as well. And so we give customers choice, et cetera. Internally, we train our Gemini models on TPUs, and we serve it that way across our products. But we use both.

Look, first of all, at a high level, NVIDIA is a phenomenal company. Jensen is awesome. We have been working with NVIDIA now for a very, very long time, and we continue to do so. And we serve a lot of the Gemini traffic on GPUs as well. And so we give customers choice, et cetera. Internally, we train our Gemini models on TPUs, and we serve it that way across our products. But we use both.

And I do think, look, I do think everyone in the industry is going to try and do something like that. But, you know, it's, you know, NVIDIA is R&D and their ability to drive that innovation. Their software stack is world class. So, you know, they have a lot of advantages as a company and I have extraordinary respect for them.

And I do think, look, I do think everyone in the industry is going to try and do something like that. But, you know, it's, you know, NVIDIA is R&D and their ability to drive that innovation. Their software stack is world class. So, you know, they have a lot of advantages as a company and I have extraordinary respect for them.

But we've always had, we are committed, we are actually deploying GPUs internally as well. I think I like that flexibility, but we are also long-term committed to the TPU direction as well. So I think it's a good combination to have both. And I think we push each other and drive the frontier forward.

But we've always had, we are committed, we are actually deploying GPUs internally as well. I think I like that flexibility, but we are also long-term committed to the TPU direction as well. So I think it's a good combination to have both. And I think we push each other and drive the frontier forward.

I think maybe it was Andrej Karpathy who used the term AJI, which is like he called it artificial jagged intelligence, right? So I think the progress is not going to be always smooth, right? Like you go through these periods, it looks like something slow, and then you see a paradigm breakthrough, et cetera. And it's been going like that for a while.

I think maybe it was Andrej Karpathy who used the term AJI, which is like he called it artificial jagged intelligence, right? So I think the progress is not going to be always smooth, right? Like you go through these periods, it looks like something slow, and then you see a paradigm breakthrough, et cetera. And it's been going like that for a while.

I think obviously over the last couple of years, you know, all of us scaled up on pre-training and then there was a lot of momentum with post-training and then with inference compute and now, you know, this progress with how do you take all that and stitch together in agentic workflows and, you know, and so on. So I do think there's a lot of progress and it feels pretty continuous to me, right?

I think obviously over the last couple of years, you know, all of us scaled up on pre-training and then there was a lot of momentum with post-training and then with inference compute and now, you know, this progress with how do you take all that and stitch together in agentic workflows and, you know, and so on. So I do think there's a lot of progress and it feels pretty continuous to me, right?

I think it's both true progress gets harder which I think will distinguish the elite teams, at least on the foundational side. I think that might be a factor. I felt the heart of the problem is, I think we are well set up for that. I think we are well set up for that. I do think we are pushing the research frontier in a much broader way than most other people.

I think it's both true progress gets harder which I think will distinguish the elite teams, at least on the foundational side. I think that might be a factor. I felt the heart of the problem is, I think we are well set up for that. I think we are well set up for that. I do think we are pushing the research frontier in a much broader way than most other people.

beyond just LLMs, transformer-based models, diffusion-based models, all those areas we are exploring in a deep, deep way. And there's always the chance that we may reach a point where you quite don't get that returns to the additional compute you're going to put in. But I quite haven't seen it yet, right? The progress looks maybe harder because you're now dealing with a lot more compute.

beyond just LLMs, transformer-based models, diffusion-based models, all those areas we are exploring in a deep, deep way. And there's always the chance that we may reach a point where you quite don't get that returns to the additional compute you're going to put in. But I quite haven't seen it yet, right? The progress looks maybe harder because you're now dealing with a lot more compute.

So you're really running into the loss of like, can I actually get as many electricians as I can to build the data centers at the speed, you know, all that stuff. But I haven't seen, or at least talking to our researchers, haven't seen anything fundamentally, hey, like we are not going to be able to move past at this point or something like that.

So you're really running into the loss of like, can I actually get as many electricians as I can to build the data centers at the speed, you know, all that stuff. But I haven't seen, or at least talking to our researchers, haven't seen anything fundamentally, hey, like we are not going to be able to move past at this point or something like that.

I think we have the opportunity to create much better experiences for people. I think people use products like Gmail, Calendar, Docs, YouTube, Search, et cetera. So with their permission, taking that personal context into account, I think we can deliver much better experiences. We are working on that, but it's something on which we have to deliver.

I think we have the opportunity to create much better experiences for people. I think people use products like Gmail, Calendar, Docs, YouTube, Search, et cetera. So with their permission, taking that personal context into account, I think we can deliver much better experiences. We are working on that, but it's something on which we have to deliver.