The Promise We Fight Every Day to Keep

At NVIDIA, we promise several things every day. One of them is that by adding a GPU into your platform and by you making the effort to program that GPU, we will accelerate your code. We will accelerate it now by a little bit. However, we want to solve for your code dramatically. The fine balance between enhancing your productivity and making it easier and easier to program this platform, while never losing sight of that delicate balance between easy to program, to use a computer, and that computer being incredibly fast - most of the applications we serve are really about speed without the speed; it is simply impossible for you to do your work. We have to find that delicate balance between programmability and speed.

One of my favorite quotes I sent this to you guys a couple of years ago was when a researcher came to me and said, "Jensen, because of your work, I'm now able to do my life's work in my lifetime." That is what it's all about. We have to find that delicate balance between programmability and speed, and access any computing architecture worth assault is really about access. You need to be able to reach out and any researcher any student any developer anywhere in the world in any industry should be able to reach out very easily and find this architecture.

One of the things that we're most proud of was a sacrifice we made early on, putting CUDA on every single GPU long before people found value in it by making it available on every single GPU from desktops to laptops to supercomputers to data centers and now on mobile devices and now in your cars. By putting CUDA in every single GPU, we make it as easy as possible for you to develop and to deploy your software. This was a strategic decision that allowed us to democratize access to our technology and unlock the full potential of computing.

Lastly, of course, the beauty of an architecture once you develop the software if your partner is dedicated to advance that architecture day in and day out day in and day out without you having to lift your finger again, the investment that you make improves over time all by itself. This is what we promised you and today I'd like to show you that in every possible way we're living up to that promise.

Ladies and gentlemen, let me introduce our brand new GPU, the Titan X. Roll it out guys so-so-so-and that's how you announce a new GPU. Ladies and gentlemen, the time next eight billion transistors; the first GPU we designed that was 4 million transistors four million transistors a small difference since then this is eight billion transistors. The team did just a fantastic job of three thousand CUDA cores, the highest single precision throughput of any GPU we've ever created is designed for single precision, is designed with a 12-gigabyte frame buffer.

And as you will see in just a moment, the reasons why we made such choices. For people who want double precision, still have Titan Z; Titan Z is still the fastest single card-belt precision GPU we have the Titan X is the highest performance single precision, the largest frame buffer, and most advanced GPU architecture - everything this is all based on Maxwell.

Ladies and gentlemen, the Titan X now let's take a look at what Titan X can do. This is work that we're about to show you that was done by a company called Epic. Epic is a long-term partner of ours; their game engine company. They are the game engine you choose if you're developing a triple A title. Now, the animation you're about to see is completely in real-time and it's running on one Titan X.

Everything you're about to see is real time running on one Titan nets it captures a hundred square miles of 3D graphics literally 100 square miles of 3D graphics basically the entire Silicon Valley you're not going to see Silicon Valley, you're going to see something a little bit more beautiful. 100 square miles; 15 million plants. You're going to see 15 million plants. This entire terrain was photometrically scanned because no it's impossible for artists to have captured landscape that beautiful. That great any given frame is 20-30 million polygons. Everything is an HDR, the camera effects are beautiful.

You're going to see physically based rendering, rocks look like rocks, water looks like water, shadows are beautiful. You might even notice some rim lining on off the shoulders. This animation this rendering in real-time graphics is the most beautiful heads that I've ever seen and I'm going out of my words yet again. Ladies and gentlemen, oh to do it's utterly utterly life changing.

And I'm going to show you some other results in just a moment but that tells you what type nexus can do and it's the reason why we do it all. Researchers all over the world 9.99 now that's right for tonight just think about that 9.99, you'll pay for it in an afternoon

"WEBVTTKind: captionsLanguage: enlet's fast forward that just a few years now there are three million cuda downloads 319 cuda applications 800 universities around the world this spans basically the globe 800 universities around the world now teach cuda and gpu acceleration 60 000 papers 60 000 papers citing the use of gpus for the research and four hundred and fifty thousand gpus now power super computers high performance computing centers all over the world and this statistics is absolutely shocking there are now 54 petabytes of combined high performance computing power that now powers the globe amazing progress amazing progress in just a few years i want to thank all of you for that the promise that we made you and the promise we fight every day to keep are several things one we promised that by adding a gpu into your platform and by you making the effort to program that gpu we will accelerate your code we will accelerate code now by a little bit we want to solve for your code dramatically number two that we seek always to find that fine balance between enhancing your productivity and making it easier and easier to program this platform enhancing your productivity while number at the same time never losing sight of that fine delicate balance between easy to program to use a computer and that computer being incredibly fast most of the applications we serve are really about speed without the speed it is simply impossible for you to do your work one of my favorite quotes and i sent this to you guys a couple of years ago was when a researcher came to me and said jensen because of your work i'm now able to do my life's work in my lifetime that is what it's all about we have to find that delicate balance between programmability and speed and number three access any computing architecture worth assault is really about access you need to be able to reach out and any researcher any student any developer anywhere in the world in any industry should be able to reach out very easily and find this architecture one of the things that we're most proud of and it was a sacrifice we made early on is put cuda on every single gpu long before people found value value in it by making it available on every single gpu from desktops to laptops to supercomputers to data centers and now on mobile devices and now in your cars by putting cuda in every single gpu we make it as easy as possible for you to develop and to deploy your software and then lastly of course the beauty of an architecture once you develop the software if your partner is dedicated to advance that architecture day in and day out day in and day out without you having to lift your finger again the investment that you make improves over time all by itself the power of moore's law and that's what we promised you and today i'd like to show you that in every possible way we're living up to that promise ladies and gentlemen let me introduce our brand new gpu the titan x roll it guys so so so and that's how you announce a new gpu ladies and gentlemen the time next eight billion transistors the first jeep the first chip we designed the first graphics chip we designed called repo 128 was 4 million transistors 4 million transistors a small difference since then this is 8 billion transistors the team did just a fantastic job three thousand cuda cores the highest single precision throughput of any gpu we've ever created is designed for single precision is designed with a 12 gigabyte frame buffer and as you will see in just a moment the reasons why we made such choices for people who want double precision we still have titan z titan z is still the fastest single card belt precision gpu we have the titan x is the highest performance single precision the largest frame buffer and most advanced gpu architecture everything this is all based on maxwell okay ladies and gentlemen the titan x now let's take a look at what titan x can do this is work that we're about to show you that was done by a company called epic epic is a long-term partner of ours their game engine company they are the game engine you choose if you're developing a triple a title now the animation you're about to see is completely in real time and it's running on one titan x everything you're about to see is real time running on one titan nets it captures a hundred square miles of 3d graphics literally 100 square miles of 3d graphics basically the entire silicon valley you're not going to see silicon valley you're going to see something a little bit more beautiful 100 square miles 15 million plants you're going to see 15 million plants this entire terrain was photometrically scanned because no it's impossible for artists to have captured landscape that beautiful that great any given frame is 20 30 million polygons everything is an hdr the camera effects are beautiful you're going to see physically based rendering the rocks look like rocks water looks like water the shadows are beautiful you might even notice some rim lining on off the shoulders this animation this rendering real-time graphics is the most beautiful heads that i've ever seen and this is just incredible work by the team at epic i love their work and they just outdone themselves yet again ladies and gentlemen oh to do it's utterly utterly life changing and i'm going to show you some other results in just a moment but that tells you what type nexus can do and it's the reason why we do it all of that all of that in your hands researchers all over the world 9.99 now that's right for tonight just think about that 9.99 you'll pay for it in an afternoonlet's fast forward that just a few years now there are three million cuda downloads 319 cuda applications 800 universities around the world this spans basically the globe 800 universities around the world now teach cuda and gpu acceleration 60 000 papers 60 000 papers citing the use of gpus for the research and four hundred and fifty thousand gpus now power super computers high performance computing centers all over the world and this statistics is absolutely shocking there are now 54 petabytes of combined high performance computing power that now powers the globe amazing progress amazing progress in just a few years i want to thank all of you for that the promise that we made you and the promise we fight every day to keep are several things one we promised that by adding a gpu into your platform and by you making the effort to program that gpu we will accelerate your code we will accelerate code now by a little bit we want to solve for your code dramatically number two that we seek always to find that fine balance between enhancing your productivity and making it easier and easier to program this platform enhancing your productivity while number at the same time never losing sight of that fine delicate balance between easy to program to use a computer and that computer being incredibly fast most of the applications we serve are really about speed without the speed it is simply impossible for you to do your work one of my favorite quotes and i sent this to you guys a couple of years ago was when a researcher came to me and said jensen because of your work i'm now able to do my life's work in my lifetime that is what it's all about we have to find that delicate balance between programmability and speed and number three access any computing architecture worth assault is really about access you need to be able to reach out and any researcher any student any developer anywhere in the world in any industry should be able to reach out very easily and find this architecture one of the things that we're most proud of and it was a sacrifice we made early on is put cuda on every single gpu long before people found value value in it by making it available on every single gpu from desktops to laptops to supercomputers to data centers and now on mobile devices and now in your cars by putting cuda in every single gpu we make it as easy as possible for you to develop and to deploy your software and then lastly of course the beauty of an architecture once you develop the software if your partner is dedicated to advance that architecture day in and day out day in and day out without you having to lift your finger again the investment that you make improves over time all by itself the power of moore's law and that's what we promised you and today i'd like to show you that in every possible way we're living up to that promise ladies and gentlemen let me introduce our brand new gpu the titan x roll it guys so so so and that's how you announce a new gpu ladies and gentlemen the time next eight billion transistors the first jeep the first chip we designed the first graphics chip we designed called repo 128 was 4 million transistors 4 million transistors a small difference since then this is 8 billion transistors the team did just a fantastic job three thousand cuda cores the highest single precision throughput of any gpu we've ever created is designed for single precision is designed with a 12 gigabyte frame buffer and as you will see in just a moment the reasons why we made such choices for people who want double precision we still have titan z titan z is still the fastest single card belt precision gpu we have the titan x is the highest performance single precision the largest frame buffer and most advanced gpu architecture everything this is all based on maxwell okay ladies and gentlemen the titan x now let's take a look at what titan x can do this is work that we're about to show you that was done by a company called epic epic is a long-term partner of ours their game engine company they are the game engine you choose if you're developing a triple a title now the animation you're about to see is completely in real time and it's running on one titan x everything you're about to see is real time running on one titan nets it captures a hundred square miles of 3d graphics literally 100 square miles of 3d graphics basically the entire silicon valley you're not going to see silicon valley you're going to see something a little bit more beautiful 100 square miles 15 million plants you're going to see 15 million plants this entire terrain was photometrically scanned because no it's impossible for artists to have captured landscape that beautiful that great any given frame is 20 30 million polygons everything is an hdr the camera effects are beautiful you're going to see physically based rendering the rocks look like rocks water looks like water the shadows are beautiful you might even notice some rim lining on off the shoulders this animation this rendering real-time graphics is the most beautiful heads that i've ever seen and this is just incredible work by the team at epic i love their work and they just outdone themselves yet again ladies and gentlemen oh to do it's utterly utterly life changing and i'm going to show you some other results in just a moment but that tells you what type nexus can do and it's the reason why we do it all of that all of that in your hands researchers all over the world 9.99 now that's right for tonight just think about that 9.99 you'll pay for it in an afternoon\n"