# (77) NVIDIA Live at CES with CEO Jensen Huang - YouTube
- Transcript ID: `0nbilspm4c4`
- Source tier: `youtube_auto`
- Original language: `en_or_source_language_unlabeled`
- Translation status: `agent_translated_reviewed` (165/165 chunks)
- Original source: https://www.youtube.com/watch?v=0NBILspM4c4
- Raw evidence: `data/transcripts/youtube/0NBILspM4c4.json`
> 中文是派生翻译，英文是留存的源文本。中文未复核时，以英文和原始音视频为准。
## 00:00:30
**中文**
准备好？去。
**English / Original**
Ready? Go.
## 00:03:57
**中文**
欢迎来到舞台，NVIDIA 创始人兼首席执行官黄仁勋。你好，拉斯维加斯。新年快乐。欢迎来到国际消费电子展。好吧，我们这里有大约 15 个主题演讲的材料。我很高兴见到你们大家。这个礼堂里有 3,000 人。院子里有 2000 人看着我们。显然还有另外 1000 人在四楼，那里本来应该是 NVIDIA 展示厅，所有人都在观看这个主题演讲，当然，世界各地有数百万人在观看
**English / Original**
Welcome to the stage, NVIDIA founder and CEO, Jensen Huang. Hello, Las Vegas. Happy New Year. Welcome to CES. Well, we have about 15 keynotes worth of material to pack in here. I'm so happy to see all of you. You got 3,000 people in this auditorium. There's 2,000 people in the courtyard watching us. There's another thousand people apparently in the fourth floor where there were supposed to be NVIDIA show floors, all watching this keynote, and of course, millions around the world are
## 00:04:38
**中文**
将观看此视频以开始新的一年。嗯，每隔 10 到 15 年，计算机行业就会重置一次。新的平台转变发生了。从大型机到PC，PC到互联网，互联网到云，云到移动。每次，应用程序世界都会瞄准一个新平台。这就是为什么它被称为平台迁移。您为新计算机编写新应用程序。
**English / Original**
going to be watching this to kick off this new year. Well, every 10 to 15 years, the computer industry resets. A new platform shift happens. From mainframe to PC, PC to internet, internet to cloud, cloud to mobile. Each time, the world of applications target a new platform. That's why it's called a platform shift. You write new applications for a new computer.
## 00:05:09
**中文**
除了这一次之外，实际上有两个同步平台迁移同时发生。虽然我们现在转向人工智能，但应用程序现在将构建在人工智能之上。起初，人们认为人工智能是应用程序，事实上，人工智能是应用程序，但你要在人工智能之上构建应用程序。但是，除此之外，您运行软件的方式、开发软件的方式也发生了根本性的变化。
**English / Original**
Except this time, there are two simultaneous platform shifts in fact happening at the same time. While we now move to AI, applications are now going to be built on top of AI. At first, people thought AIs are applications, and in fact, AIs are applications, but you're going to build applications on top of AIs. But, in addition to that, how you run the software, how you develop the software fundamentally changed.
## 00:05:44
**中文**
计算机行业的整个五层堆栈正在被重塑。您不再对软件进行编程，而是训练软件。你不是在 CPU 上运行它，而是在 GPU 上运行它。虽然应用程序是预先录制、预先编译并在您的设备上运行的，但现在应用程序可以理解上下文并每次都完全从头开始生成每个像素、每个 token。
**English / Original**
The entire five-layer stack of the computer industry is being reinvented. You no longer program the software, you train the software. You don't run it on CPUs, you run it on GPUs. And whereas applications were pre-recorded, pre-compiled, and run on your device, now applications understand the context and generate every single pixel, every single token, completely from scratch every single time.
## 00:06:17
**中文**
由于加速计算和人工智能，计算已经从根本上重塑。五层蛋糕的每一层现在都在重新设计。嗯，这意味着过去 10 年大约 10 万亿美元的计算现在正在现代化，以适应这种新的计算方式。这意味着数千亿美元，每年数千亿美元的风险投资，将用于现代化和发明这个新世界。这意味着一百万亿美元的工业（其中百分之几是研发预算）正在转向人工智能。
**English / Original**
Computing has been fundamentally reshaped as result of accelerated computing, as result of artificial intelligence. Every single layer of that five-layer cake is now being reinvented. Well, what that means is some 10 trillion dollars or so of the last decade of computing is now being modernized to this new way of doing computing. What that means is hundreds of billions of dollars, a couple hundred billion dollars in VC funding each year, is going into modernizing and inventing this new world. And what it means is a hundred trillion dollars of industry, several percent of which is R&D budget, is shifting over to artificial intelligence.
## 00:07:00
**中文**
人们问：“钱从哪里来？”这就是钱的来源。人工智能向人工智能的现代化，研发预算从经典方法转向现在的人工智能方法。大量投资进入这个行业，这解释了为什么我们如此忙碌。去年也没有什么不同。
**English / Original**
People ask, "Where is the money coming from?" That's where the money is coming from. The modernization of AI to AI, the shifting of R&D budgets from classical methods to now artificial intelligence methods. Enormous amounts of investments coming into this industry, which explains why we're so busy. And this last year was no difference.
## 00:07:21
**中文**
去年真是令人难以置信。去年，有一张幻灯片即将推出。当你不练习时就会发生这种情况。 2025 年是令人难以置信的一年。只是看起来一切似乎都在同时发生，事实上，很可能是这样。当然，第一件事是缩放定律。 2015 年，我认为真正会产生影响的第一个语言模型产生了巨大的影响。它被称为 BERT。 2017年，Transformer来了。
**English / Original**
This last year was incredible. This last year, there's a slide coming. This is what happens when you don't practice. 2025 was an incredible year. It's just see It seems like everything was happening all at the same time, and in fact, it probably was. The first thing, of course, is scaling laws. In 20 15, the first language model that I thought was really going to make a difference made a huge difference. It was called BERT. 2017, transformers came.
## 00:07:56
**中文**
直到 5 年后的 2022 年，ChatGPT 时刻的到来，才唤醒了世界人工智能的可能性。一年后发生了一件非常重要的事情。 ChatGPT的首个 o1 模型，第一个推理模型，完全革命性的，发明了这个想法，叫做测试时扩展，这是非常常识常识的事情。我们不仅预先训练模型进行学习，还通过强化学习对其进行后训练，以便它能够学习技能。
**English / Original**
It wasn't until 5 years later, 2022, that ChatGPT moment happened, and it awakened the world to the possibilities of artificial intelligence. Something very important happened a year after that. The first 01 model from ChatGPT, the first reasoning model, completely revolutionary, invented this idea called test time scaling, which is very common sense common sensical thing. Not only do we pre-train a model to learn, we post-train it with our reinforcement learning so that it could learn skills.
## 00:08:29
**中文**
现在我们还有测试时扩展，这是另一种思考方式。你实时思考。人工智能的每一阶段都需要大量的计算。计算法则继续扩展。大型语言模型继续变得更好。与此同时，另一个突破发生了。而这一突破发生在2024年。
**English / Original**
And now we also have test time scaling, which is another way of saying thinking. You think in real time. Each one of these phases of artificial intelligence requires enormous amount of compute. And the computing law continued to scale. Large language models continued to get better. Meanwhile, another breakthrough happened. And this breakthrough happened in 2024.
## 00:08:52
**中文**
智能体系统开始出现，到 2025 年，它开始渗透到几乎所有地方。智能体模型具有推理、查找信息、研究、使用工具、计划未来、模拟结果的能力，突然开始解决非常非常重要的问题。我最喜欢的智能体模型之一称为 Cursor，它彻底改变了我们在 NVIDIA 进行软件编程的方式。智能体系统将从这里真正起飞。
**English / Original**
Agentic systems starting to emerge, and 2025, it started to pervade to to proliferate just about everywhere. Agentic models that have the ability to reason, look up information, do research, use tools, plan futures, simulate outcomes, all of a sudden started to solve very, very important problems. One of my favorite agentic models is called Cursor, which revolutionized the way we do software programming at NVIDIA. Agentic systems are going to really take off from here.
## 00:09:26
**中文**
当然，还有其他类型的人工智能。我们知道大型语言模型并不是唯一的信息类型。只要宇宙有信息，只要宇宙有结构，我们就可以教授一个大型语言模型，一种语言模型的形式，去理解该信息，理解其表示，并将其转化为人工智能。最大、最重要的人工智能之一是物理 AI，即理解自然法则的人工智能。
**English / Original**
Of course, there were other types of AI. We know that large language models isn't the only type of information. Wherever the universe has information, wherever the universe has structure, we could teach a large language model, a form of language model, to go understand that information, to understand its representation, and to turn that into an AI. One of the biggest, most important one is physical AI, AIs that understand the laws of nature.
## 00:09:55
**中文**
当然，物理 AI是指人工智能与世界互动。但是，世界本身就有信息编码的信息，这就是所谓的AI 物理。人工智能，就物理 AI而言，你有与物理世界交互的人工智能，你有AI 物理，人工智能理解物理定律。最后，去年发生的最重要的事情之一是开放模型的进步。
**English / Original**
And then, of course, physical AIs is about AIs interacting with the world. But, the world itself has information encoded information, and that's called AI physics. AI that in the case of physical AI, you have AI that interacts with the physical world, and you have AI physics, AI that understands the laws of physics. And then, lastly, one of the most important things that happened last year, the advancement of open models.
## 00:10:23
**中文**
我们现在可以知道，当开源、开放创新、全球每家公司和每一个行业的创新同时被激活时，人工智能将在各地蓬勃发展。开放模型在去年才真正起飞。事实上，去年我们看到了 DeepSeek R1 的进步，这是第一个开放模型推理系统。
**English / Original**
We can now know that AI is going to proliferate everywhere when open source, when open innovation, when innovation across every single company and every industry around the world is activated at the same time. Open models really took off last year. In fact, last year, we saw the advance of DeepSeek R1, the first open model that's a reasoning system.
## 00:10:50
**中文**
它让世界大吃一惊，并激活了整个运动。真的非常令人兴奋的工作。我们对此感到非常高兴。现在，我们在世界各地开放了各种不同类型的开放模型系统，而且我们现在知道开放模型也已经达到了前沿。仍然比前沿模型落后 6 个月，但每 6 个月就会出现一个新模型，而且这些模型变得越来越智能。
**English / Original**
It caught the world by surprise, and it activated literally this entire movement. Really, really exciting work. We're so happy with it. Now, we have openings open model systems all over the world of all different kinds, and we now know that open models have also reached the frontier. Still solidly is 6 months behind the frontier models, but every single 6 months a new model is emerging, and these models are getting smarter and smarter.
## 00:11:22
**中文**
因此，您可以看到下载数量呈爆炸式增长。下载量增长如此之快，是因为初创公司想要参与人工智能革命，大公司想要参与，研究人员想要参与，学生想要参与，几乎每个国家都想要参与。智能，智能的数字形式，怎么可能会把任何人抛在后面呢？因此，开放模型去年确实彻底改变了人工智能。整个行业将因此而重塑。现在，我们不久前就有了这样的预感。
**English / Original**
Because of that, you could see the number of downloads has exploded. The number of downloads is growing so fast because startups want to participate in the AI revolution, large companies want to, researchers want to, students want to, just about every single country wants to. How is it possible that intelligence, the digital form of intelligence, will leave anyone behind? And so, open models has really revolutionized artificial intelligence last year. This entire industry is going to be reshaped as a result of that. Now, we had this inkling some time ago.
## 00:11:59
**中文**
您可能听说过，几年前，我们刚刚开始构建和运营自己的人工智能超级计算机。我们称之为 DGX 云。很多人问，“你要进军云业务吗？”答案是否定的。我们正在构建这些 DGX 超级计算机供我们自己使用。事实证明，我们有数十亿美元的超级计算机正在运行，这样我们就可以开发我们的开放模型。我对我们正在做的工作非常满意。
**English / Original**
You might have heard that several years ago, we just started to build and operate our own AI supercomputers. We call them DGX Clouds. A lot of people asked, "Are you going to going into the cloud business?" The answer is no. We're building these DGX supercomputers for our own use. Well, it turns out we have billions of dollars of supercomputers in operation, so that we could develop our open models. I am so pleased with the work that we're doing.
## 00:12:28
**中文**
它开始引起全世界和各行业的关注，因为我们正在许多不同领域进行前沿人工智能模型工作。我们在蛋白质、数字生物学、La-Proteina 领域所做的工作是为了能够合成和生成蛋白质。 OpenFold 3 了解蛋白质的结构。
**English / Original**
It is starting to attract attention all over the world and all over the industries, because we are doing frontier AI model work in so many different domains. The work that we did in proteins, in digital biology, La Pertina, to be able to synthesize and generate proteins. OpenFold 3 to understand the understand the structure of proteins.
## 00:12:49
**中文**
Evo 2，如何理解和生成多种蛋白质，否则是细胞细胞表征的开始。地球2，理解物理定律的人工智能。我们与 FourCastNet 所做的工作，与 CorrDiff 所做的工作，确实彻底改变了人们进行天气预报的方式。 NeMotron，我们现在正在那里进行开创性的工作。第一个混合变压器 SSM 模型速度快得令人难以置信，因此可以思考很长时间，或者可以非常快速地思考不太长的时间，并产生非常非常聪明的智能答案。 NeMotron 3 是一项开创性的工作，您可以期待
**English / Original**
Evo 2, how to understand and generate multiple proteins, otherwise the beginnings of cellular cellular representation. Earth 2, AI that understands the laws of physics. The work that we did with ForecastNet, the work that we did with ClimaDiv, really revolutionized the way that people are doing weather prediction. NeMo-Megatron, we've now doing groundbreaking work there. The first hybrid transformer SSM model that's incredibly fast and therefore can think for a very long time, or can think very quickly with that for not a very long time, and produce very very smart intelligent answers. NeMo-Megatron 3 is groundbreaking work, and you can expect
## 00:13:32
**中文**
我们将在不久的将来提供 NeMotron 3 的其他版本。 Cosmos，一种前沿开放世界基础模型，了解世界如何运作。 GR00T，一个人形机器人系统，关节、移动性、运动，这些模型、这些技术现在正在被整合，并且在每一个案例中都向世界开放。向世界开放的前沿人形机器人模型。然后，今天，我们要谈谈 Alpamayo，我们在自动驾驶汽车方面所做的工作。我们不仅开源模型，还开源用于训练这些模型的数据
**English / Original**
us to deliver other versions of NeMo-Megatron 3 in the near future. Cosmo, a frontier open world foundation model, one that understand how the world works. Groot, a humanoid robotic system, articulation, mobility, locomotion, these models, these technologies are now being integrated, and in the each one of these cases open to the world. Frontier humanoid robotics models open to the world. And then, today, we're going to talk a little bit about ElpaMayo, the work that we've been doing in self-driving cars. Not only do we open source the models, we also open source the data that we used to train those
## 00:14:14
**中文**
模型。
**English / Original**
models.
## 00:14:15
**中文**
因为只有这样，你才能真正相信模型是如何形成的。我们开源所有模型。我们帮助您利用它们制作衍生品。我们有一整套库。我们将它们称为 NeMo 库、物理 NeMo 库和 Clara NeMo 库，以及生物 NeMo 库。这些库中的每一个都是人工智能的生命周期管理系统，因此您可以处理数据，可以生成数据，可以训练模型，可以创建模型，评估模型，保护模型，一直到部署模型。
**English / Original**
Because that in that way, only in that way can you truly trust how the models came to be. We open source all the models. We help you make derivatives from them. We have a whole suite of libraries. We call them NeMo libraries, physics NeMo libraries, and the Clara NeMo libraries, each bio NeMo libraries. Each one of these libraries are life cycle management systems of AIs, so that you could process the data, you could generate data, you could train the model, you could create the model, evaluate the model, guardrail the model, all the way to deploying the model.
## 00:14:47
**中文**
这些库中的每一个都非常复杂，并且全部都是开源的。因此，现在，在这个平台之上，NVIDIA 是一家前沿的人工智能模型构建者，我们以一种非常特殊的方式构建它。我们完全开放地构建它，以便我们能够使每个公司、每个行业、每个国家都成为这场人工智能革命的一部分。我对我们在那里所做的工作感到无比自豪。事实上，如果您注意到这些图表，就会发现我们对这个行业的贡献是无可比拟的，事实上，您会看到我们将继续这样做，甚至加速。
**English / Original**
Each one of these libraries are incredibly complex, and all of it is open sourced. And so, now, on top of this platform, Nvidia is a frontier AI model builder, and we build it in a very special way. We build it completely in the open, so that we can enable every company, every industry, every country to be part of this AI revolution. I'm incredibly proud of the work that we're doing there. In fact, if you notice the the charts, the chart shows that our contribution to this industry is bar none, and you're going to see us, in fact, continue to do that, if not accelerate.
## 00:15:26
**中文**
这些模型也是世界一流的。它们不仅开放，而且还在排行榜上名列前茅。这是我们非常自豪的领域。他们在智力排行榜上名列前茅。呃，我们有呃重要的模型，可以理解多模态文档，也称为 PDF。世界上最有价值的内容都集中在 PDF 中。但是，它们需要人工智能来找出里面的内容，解释里面的内容，并帮助你阅读它。因此，我们的 PDF 检索器、PDF 解析器是世界一流的。
**English / Original**
These models are also world-class. Not only are they open, they're also top the leaderboards. This is an area where we're very proud. They top leaderboards in intelligence. Uh we have uh important models that are understand multi-modality documents, otherwise known as PDFs. The most valuable content in the world are captured in PDFs. But, they're it takes artificial intelligence to find out what's inside, interpret what's inside, and help you read it. And so, our PDF retrievers, our PDF parsers, are world-class.
## 00:15:59
**中文**
我们的语音识别模型，绝对是世界一流的。我们的检索模型，基本上是搜索语义搜索AI搜索，现代AI时代的数据库引擎，世界一流。因此，我们始终位居排行榜前列。这是我们非常自豪的领域。所有这些都是为了您构建AI 智能体的能力。
**English / Original**
Our speech recognition models, absolutely world-class. Our retrieval models, basically search semantic search AI search, the database engine of the modern AI era, world-class. So, we're on top of leaderboards constantly. This is an area we're very proud of. And all of that is in service of your ability to build AI agents.
## 00:16:24
**中文**
这确实是一个突破性的发展领域。你知道，我第一次当 ChatGPT 出现时，人们说，你知道，天哪，它产生了非常有趣的结果，但它产生了很大的幻觉。而它之所以会产生幻觉,当然是它可以记住嗯过去的一切,但是现在却无法记住未来的一切。
**English / Original**
This is really a groundbreaking area of development. You know, I first when people when ChatGPT came out, people said, you know, uh gosh, it it it produced really interesting results, but it hallucinated greatly. And the reason why it hallucinated, of course, it could memorize everything um in the past, but it can't memorize everything in the future in the current.
## 00:16:43
**中文**
因此，它需要以研究为基础。在回答问题之前，它必须先进行基础研究。推理能力，我必须做研究吗？我必须使用工具吗？如何将问题分解为步骤。这些步骤中的每一个都是人工智能模型知道如何做的事情，并且它能够将其组合成一系列步骤来执行以前从未做过、从未接受过训练的事情。
**English / Original**
And so, it needs to be grounded in research. It has to do fundamental research before it answers a question. The ability to reason about, do I have to do research? Do I have to use tools? How do I break up a problem into steps. Each one of these steps something that that the AI model knows how to do, and together it is able to compose it into a sequence of steps to perform something it's never done before, it's never been trained to do.
## 00:17:09
**中文**
这就是推理的奇妙能力。我们可以遇到我们以前从未见过的情况，并将其分解为我们知道如何做的情况和知识或规则，因为我们过去经历过。因此，人工智能模型现在的推理能力非常强大。
**English / Original**
This is the wonderful capability of reasoning. We could we could be we can encounter a circumstance we've never seen before, and break it down into circumstances and knowledge or rules that we know how to do, because we've experienced it in past. And so, the ability for AI models now to be able to reason incredibly powerful.
## 00:17:30
**中文**
智能体的推理能力为所有这些不同的应用打开了大门。我们不再需要训练人工智能模型在第一天就知道一切。就像我们不必在第一天就知道一切一样。我们应该能够在任何情况下推理如何解决该问题。大型语言模型现在已经实现了这一根本性的飞跃。使用强化学习和思维链的能力，你知道，搜索和规划以及所有这些不同的技术和强化学习使我们能够拥有这种基本能力，而且现在它也是完全开源的。但是，真正了不起的是发生的另一个突破。
**English / Original**
The reasoning capability of agents opened the doors to all of these different applications. We no longer have to train an AI model to know everything on day one. Just as we don't have to know everything on day one. That we should be able to in every circumstance reason about how to solve that problem. Large language models has now made this fundamental leap. The ability to use reinforcement learning and chain of thoughts and you know, search and planning and all these different techniques and reinforcement learning has made it possible for us to have this basic capability, and it's also now completely open sourced. But, the thing that's really terrific is another breakthrough that happened.
## 00:18:09
**中文**
我第一次看到它是在 Eran 的 Perplexity，Perplexity，搜索公司，人工智能搜索公司，非常非常创新的公司。当我第一次意识到他们同时使用多个模型时，我认为这完全是天才。当然，我们会这样做。当然，人工智能也会呼吁世界上所有伟大的人工智能在推理链的任何部分解决它想要解决的问题。这就是人工智能真正是多模态的原因，这意味着它们理解语音、图像、文本、视频和 3D
**English / Original**
And the first time I saw it was with Eran's Perplexity, Perplexity, the search company, the AI search company, really really innovative company. And the first time I realized they were using multiple models at the same time, I thought it was completely genius. Of course, we would do that. Of course, an AI would also call upon all of the world's great AIs to solve the problem it wants to solve at any part of the reasoning chain. And this is the reason why AIs are really multi- modal, meaning they understand speech and images and text and videos and 3D
## 00:18:50
**中文**
图形和蛋白质。它是多模态的。
**English / Original**
graphics and proteins. It's multi-modal.
## 00:18:52
**中文**
它也是多模型的，这意味着它应该能够使用最适合任务的任何模型。因此，从定义上来说，它是多云，因为这些人工智能模型位于所有这些不同的地方，而且它也是混合云。因为如果您是一家企业公司，或者您制造了一个机器人，或者无论该设备是什么，有时它位于边缘，有时位于无线电信号塔，有时可能位于企业中，或者可能位于医院的地方，您需要在您身边实时获取数据。无论这些应用程序是什么，我们现在知道这就是人工智能应用程序
**English / Original**
It's also multi-model, meaning that it should be able to use any model that best fits the task. It is multi-cloud by definition, therefore, because these AI models are sitting in all these different places, and it also is hybrid cloud. Because if you're an enterprise company, or you built a robot, or whatever that device is, sometimes it's at the edge, sometimes a radio cell tower, maybe sometimes it's in an enterprise, or maybe it's a place where a hospital where you need to have the the data in real time right next to you. Whatever those applications are, we know now this is what an AI application
## 00:19:34
**中文**
看起来像在未来。
**English / Original**
looks like in the future.
## 00:19:36
**中文**
或者换个角度思考，因为未来的应用是建立在人工智能之上的，这就是未来应用的基本框架。这个基本框架，这个智能体 AI的基本结构，可以完成我正在谈论的事情。多模型现在已经推动了各种人工智能初创公司的发展。现在，由于我们为您提供了所有开放模型和所有工具，您还可以自定义您的人工智能来教授其他人没有教授的人工智能技能。没有其他人能够以这种方式让他们的人工智能变得聪明或聪明。你可以这样做
**English / Original**
Or another way to think about that, because future applications are built on AIs, this is the basic framework of future applications. This basic framework, this basic structure of agentic AIs that could do the things that I'm talking about. That is multi-model has now turbocharged AI startups of all kinds. And now you can also because of the all of the open models and all the tools that we've provided you, you could also customize your AIs to teach your AI skills that nobody else is teaching. Nobody else is causing their AI to become intelligent or smart in that way. You could do it for
## 00:20:19
**中文**
你自己。这就是我们对 NeMotron、NeMo 所做的工作，以及我们对开放模型所做的所有事情的目的。你在它前面放置一个智能路由器，该路由器本质上是一个管理器，它根据你给出的提示的意图来决定哪一个任务，哪一个模型最适合该应用程序，以解决该问题。
**English / Original**
yourself. And that's the work that we do with Nemotron, Nemo, and all of the things that we do with open models is intended to do. You put a smart router in front of it and that router is essentially a manager that decides which one of the task, based on the intention of the prompts that you give it, which one of the models is best fit for that application, for that solving that problem.
## 00:20:41
**中文**
好的？那么现在当你思考这个架构时，你有什么？当你想到这个架构时，突然间你就拥有了一个人工智能，它一方面可以完全由你定制。您可以教授一些东西，让您为公司发挥自己的技能。域秘密的东西。您拥有深厚的领域专业知识的东西。也许您已经获得了训练该人工智能模型所需的所有数据。另一方面，根据定义，你的人工智能始终处于前沿。
**English / Original**
Okay? So now when you think about this architecture, what do you have? When you think about this architecture, all of a sudden you have an AI that's on the one hand completely customizable by you. Something that you could teach to do your own very skills for your company. Something that's domain secret. Something where you have deep domain expertise. Maybe you've got all of the data that you need to train that AI model. On the other hand, your AI is always at the frontier by definition.
## 00:21:15
**中文**
一方面，您始终处于前沿，另一方面，您始终处于定制状态，并且它应该运行。因此，我们认为我们应该制作最简单的示例以供您使用。我们将整个框架称为蓝图。我们的蓝图已集成到世界各地的企业 SaaS 平台中，我们对进展感到非常满意。但我们所做的是向您展示一个任何人都可以做的事情的简短示例。
**English / Original**
You're always at the frontier on the one hand, you're always customized on the other hand, and it should just run. And so we thought we would make the simplest of examples to make it available to you. This entire framework we call a blueprint. And we have blueprints that are integrated into enterprise SaaS platforms all over the world and we're really pleased with the progress. But what we do is show you a a short example of something that anybody can do.
## 00:21:44
**中文**
让我们建立一个私人助理。我希望它能帮助我处理日历、电子邮件、待办事项列表，甚至监视我的家。我使用 Brev 将我的 DGX Spark 变成个人云。因此，无论我使用云 GPU 还是 DGX Spark，我都可以使用相同的接口。我使用前沿模型 API 来轻松入门。我希望它能帮助我处理电子邮件，因此我创建了一个电子邮件工具供我的智能体拨打电话。
**English / Original**
Let's build a personal assistant. I want it to help me with my calendar, emails, to-do lists, and even keep an eye on my home. I use Brev to turn my DGX Spark into a personal cloud. So I can use the same interface whether I'm using a cloud GPU or a DGX Spark. I use a frontier model API to easily get started. I want it to help me with my emails, so I create an email tool for my agent to call.
## 00:22:13
**中文**
我希望我的电子邮件保持私密性，因此我将添加一个在 Spark 本地运行的开放模型。现在，对于任何工作，我希望智能体使用正确的模型来完成正确的任务，因此我将使用基于意图的模型路由器。这样，需要电子邮件的提示将保留在我的 Spark 上，而其他一切都可以调用前沿模型。我希望我的助手能够与我的世界互动，因此我会将其连接到 Hugging Face 的 Reachy 迷你机器人。
**English / Original**
I want my emails to stay private, so I'll add an open model that's running locally on the Spark. Now for any job, I want the agent to use the right model for the right task, so I'll use an intent-based model router. This way, prompts that need email will stay on my Spark and everything else can call the frontier model. I want my assistant to interact with my world, so I'll hook it up to Hugging Face's Reachy mini robot.
## 00:22:43
**中文**
我的智能体通过工具调用控制 Reachy 的头部、耳朵和摄像头。我想给 Reachy 一个声音，而且我真的很喜欢 11 Labs，所以我会连接他们的 API。大家好，我是 Reachy，在 DGX Spark 上运行。嘿，Reachy，我今天的待办事项清单上有什么？你今天的待办事项清单。拿上杂货、鸡蛋、牛奶、黄油，然后把新剧本寄给詹森。
**English / Original**
My agent controls the head, ears, and camera of the Reachy with tool calls. I want to give Reachy a voice and I really like 11 Labs, so I'll hook up their API. Hi, I'm Reachy, running on DGX Spark. Hey Reachy, what's on my to-do list today? Your to-do list today. Grab groceries, eggs, milk, butter, and send Jensen the new script.
## 00:23:07
**中文**
好的，让我们向 Jensen 发送更新。告诉他我们会在今天结束之前为他准备好。会做。雷奇，还有一张草图。你能把它变成建筑效果图吗？当然。好的。现在制作一个视频并带我参观房间。干得好。那太棒了。借助 Brev，我可以共享对 Spark 和 Reachy 的访问权限，因此我将与 Anna 共享它。
**English / Original**
Okay, let's send Jensen that update. Tell him we'll have it for him by the end of day. Will do. Reachy, there's a sketch too. Can you turn it into an architectural rendering? Sure. Nice. Now make a video and show me around the room. Here you go. That's great. With Brev, I can share access to my Spark and Reachy, so I'm going to share it with Anna.
## 00:23:42
**中文**
嘿，雷奇，土豆在做什么？他在沙发上。我记得你不喜欢这个。我会告诉他下车。土豆，从沙发上下来。随着开源领域的不断进步，您可以构建出令人难以置信的东西。我很想看看你创造了什么。这不是不可思议吗？令人惊奇的是，现在这一切都变得微不足道了。
**English / Original**
Hey Reachy, what's Potato up to? He's on the couch. I remember you don't like this. I'll tell him to get off. Potato, off the couch. With all the progress in open source, it's incredible to see what you can build. I'd love to see what you create. Isn't that incredible? Now the amazing thing is that is utterly trivial now.
## 00:24:11
**中文**
现在这完全是微不足道的。然而就在几年前，这一切都是不可能的，绝对不可想象的。嗯，这个基本框架，这种使用预先训练的语言模型构建应用程序的基本方法，它们是专有的，它们是前沿的，将其与定制的语言模型结合成一个智能体框架，一个推理框架，允许您访问工具和文件，甚至可以连接到其他智能体。
**English / Original**
That is utterly trivial now. And yet just a couple years ago all of that would have been impossible, absolutely unimaginable. Well, this basic framework, this basic way of building applications using language models that are pretrained and they're proprietary, they're frontier, combine it with customized language models into a agentic framework, a reasoning framework that allows you to access tools and files and maybe even connect to other agents.
## 00:24:43
**中文**
这基本上就是现代人工智能应用程序或应用程序的架构。我们创建这些应用程序的速度非常快。请注意，如果您向该应用程序提供它以前从未见过的信息，或者其结构与您想象的不完全一致，它仍然可以通过它进行推理，并尽最大努力通过数据和信息进行推理，以尝试了解如何解决问题。
**English / Original**
This is basically the architecture of AI applications or applications in the modern age. And the ability for us to create these applications are incredibly fast. And notice if you give it this application um information that it's never seen before or in a structure that has is not represented exactly as you thought, it can still reason through it and make its best effort to reason through the data, the information to try to understand how to solve the problem.
## 00:25:18
**中文**
人工智能。好的，现在正在集成这个基本框架，以及我刚才描述的所有内容，我们受益于与一些世界领先的企业平台公司的合作。呃，帕兰提尔，例如。我嗯他们的整个人工智能和数据处理平台今天正在由 NVIDIA 集成和加速。
**English / Original**
Artificial intelligence. Okay, so this basic framework is now being integrated and and everything that I just described, we have the benefit of working with some of the world's leading enterprise platform companies. Uh Palantir, for example. I um their their entire AI and data processing platform is being integrated and accelerated by NVIDIA today.
## 00:25:38
**中文**
ServiceNow，全球领先的客户服务和员工服务平台。 Snowflake，全球顶级的云数据平台。呃，令人难以置信的工作正在那里进行。嗯，代码兔。呃，我们在 NVIDIA 上都使用 Code Rabbit。呃 CrowdStrike 创建 AI 来检测、定义 AI 威胁。嗯，NetApp，他们的人工智能数据平台现在拥有 NVIDIA 语义人工智能和智能体系统，以便他们提供客户服务。但重要的是，这不仅是您现在开发应用程序的方式，而且还将成为您平台的用户界面。那么无论是
**English / Original**
ServiceNow, the world's leading customer service and um employee service platform. Snowflake, the world's top data platform in the cloud. Uh incredible work that that uh is being done there. Uh Code Rabbit. Uh we're using Code Rabbit all over NVIDIA. Uh CrowdStrike creating AIs to detect, define AI threats. Uh NetApp, their AI their data platform now has NVIDIA semantic AI on top of it and agentic systems on top of it for them to do customer service. But the important thing is this, not only is this the way that you develop applications now, this is going to be the user interface of your platform. So whether it's
## 00:26:20
**中文**
Palantir、ServiceNow 或 Snowflake 以及我们正在合作的许多其他公司，智能体系统就是界面。它不再是带有一堆需要您输入信息的方块的 Excel。也许不再只能使用命令行了。现在，所有这些多模态信息都是可能的，并且您与平台的交互方式变得更好，如果您愿意的话，很简单，就像您与人交互一样。
**English / Original**
Palantir or ServiceNow or Snowflake and many other companies that we're working with, the agentic system is the interface. It's no longer Excel with a bunch of, you know, squares that you enter enter information into. Maybe it's no longer could just command line. The any all of that multimodality information is now possible and the way you interact with your platform is much more well, if you will, simple, like you're interacting with people.
## 00:26:50
**中文**
这就是智能体系统正在彻底改变企业人工智能。接下来是物理 AI。这是我多年来谈论的一个领域。事实上，我们为此已经努力了八年。问题是，如何将计算机内部通过屏幕和扬声器与你交互的智能设备转变为可以与世界互动的设备，这意味着它可以理解世界如何运作的常识。
**English / Original**
And so that's enterprise AI being revolutionized by agentic systems. The next thing is physical AI. This is an area that you've seen me talk about for several years. In fact, we've been working on this for eight years. The question is, how do you take something that is intelligent inside a computer and interacts with you with screens and speakers to something that can interact with the world, meaning it can understand the common sense of how the world works.
## 00:27:22
**中文**
物体的持久性。如果我移开视线再回头看，那个物体仍然在那里。嗯因果关系。如果我推它，它就会翻倒。它理解摩擦力和重力。它理解惯性。一辆在路上行驶的重型卡车需要多一点时间才能停下来。一个球将继续滚动。这些想法对于小孩子来说都是常识，但对于人工智能来说却是完全未知的。因此，我们必须创建一个系统，让人工智能能够学习物理世界的常识，学习其规律，当然也能够从中学习
**English / Original**
Object permanence. If I look away and I look back, that object is still there. Um causality. If I push it, it tips over. It understands friction and gravity. It understands inertia. That a heavy truck rolling down the road is going to need a little bit more time to stop. That a ball is going to keep on rolling. These ideas are common sense to even a little child, but for AI, it's completely unknown. And so we have to create a system that allows AIs to learn the the common sense of the physical world, learn its laws, but also to be able to, of course, learn from
## 00:28:03
**中文**
数据非常稀缺，要能够评估人工智能是否有效，这意味着它必须在环境中进行模拟。如果人工智能没有能力模拟物理世界对其行为的反应，那么它如何知道它正在执行的操作与它应该做的一致？它的动作的响应对于模拟来说非常重要，否则无法评估它，每次都不一样。
**English / Original**
data and the data is quite scarce, and to be able to evaluate whether that AI is working, meaning it has to simulate in an environment. How does an AI know that the the actions that it's performing is consistent with what it should do if it doesn't have the ability to simulate the response of the physical world back on its actions? The response of its actions is really important to simulate, otherwise there's no way to evaluate it, it's different every time.
## 00:28:31
**中文**
所以这个基本系统需要三台计算机。当然，我们知道 NVIDIA 为训练 AI 模型而构建的一台计算机。我们知道的另一台计算机是对模型进行计算推理。推理模型本质上是一台机器人计算机，可以在汽车中运行，或者在机器人中运行，或者在工厂中运行，在边缘的任何地方运行。
**English / Original**
And so this basic system requires three computers. One computer, of course, the one that we know that NVIDIA builds for training the AI models. Another computer that we know is to inference the compute inference the models. Inferencing the model is essentially a robotics computer that runs in the car or runs in a robot or runs in a factory, runs anywhere at the edge.
## 00:28:55
**中文**
但必须有另一台专为模拟而设计的计算机。模拟几乎是 NVIDIA 所做一切的核心。这是我们最舒服的地方。模拟实际上是我们用物理 AI所做的几乎所有事情的基础。所以我们有三台计算机和在这些计算机上运行的多个堆栈，这些库使它们变得有用。 Omniverse 是我们基于物理的数字孪生模拟世界。
**English / Original**
But there has to be another computer that's designed for simulation. And simulation is at the heart of almost everything NVIDIA does. This is This is where we are most comfortable. And simulation was really the foundations of almost everything that we've done with physical AI. So we have three computers and multiple stacks that run on these computers, these libraries to make them useful. Omniverse is our digital twin physically based simulation world.
## 00:29:24
**中文**
正如我前面提到的，Cosmos 是我们的基础模型，不是语言的基础模型，而是世界的基础模型。而且它也与语言保持一致。你可以说，你知道，球发生了什么？他们会告诉你球正在街上滚。因此，世界基础模型，当然还有机器人模型。我们有两个。其中之一叫GR00T。
**English / Original**
Cosmos, as I mentioned earlier, is our foundation model, not a foundation model for language, but a foundation model of the world. And it's also aligned with language. You could say something like, you know, what's happening to the ball? And they'll tell you the ball's rolling down the street. And so a world foundation model and then, of course, the robotics models. We have two of them. One of them is called Groot.
## 00:29:48
**中文**
另一个叫做 Alpamayo，我将向您介绍它。现在，我们对物理 AI要做的最重要的事情之一就是首先创建数据来训练人工智能。这些数据从哪里来？我们不是因为创建了一堆我们认为人工智能可以从中学习的基本事实而拥有语言，而是如何教AI 物理的基本事实？有很多很多的视频，很多很多的视频，但不足以捕捉我们所需的多样性和互动类型。
**English / Original**
The other one's called Alpha Mylo that I'm going to tell you about. Now, the one of the most important things that we have to do with physical AI is to create the data to train the AI in the first place. Where does that data come from? Rather than instead of having languages because we created a bunch of text that are what we consider ground truth that the AI can learn from, how do we teach an AI the ground truth of physics? There are lots and lots of videos, lots and lots of videos, but hardly enough to capture the diversity and the type of interactions that we need.
## 00:30:21
**中文**
因此，这就是伟大思想聚集在一起并将过去的计算转化为数据的地方。现在，利用以物理定律为基础和条件的合成数据生成，以事实真相为基础和条件，我们现在可以有选择地巧妙地生成数据，然后用这些数据来训练人工智能。因此，例如，进入这个人工智能，左边的这个 Cosmos 人工智能世界模型，是交通模拟器的输出。
**English / Original**
And so, this is where great minds came together and transformed what used to be compute into data. Now, using synthetic data generation that is grounded and conditioned by the laws of physics, grounded and conditioned by ground truth, we can now selectively cleverly generate data that we can then use to train the AI. So, for example, what comes into this AI, this Cosmos AI world model on the left on over here, is the output of a traffic simulator.
## 00:31:01
**中文**
现在，这个交通模拟器还不足以让人工智能学习。我们可以将其放入 Cosmos 基础模型中，并生成基于物理且物理合理的环绕视频，人工智能现在可以从中学习。这方面的例子有很多。让我向您展示 Cosmos 可以做什么。物理 AI 的ChatGPT 时刻即将到来。
**English / Original**
Now, this traffic simulator is hardly enough for an AI to learn from. We can take this, put it into a Cosmos Foundation model, and generate surround video that is physically based and physically plausible that the AI can now learn from. And there are so many examples of this. Let me show you what Cosmos can do. The Chat GPT moment for physical AI is nearly here.
## 00:31:32
**中文**
但挑战是显而易见的。物理世界是多种多样且不可预测的。收集真实世界的训练数据既缓慢又昂贵，而且还远远不够。答案是合成数据。这一过程始于 NVIDIA Cosmos，这是一个物理 AI 的开放前沿世界基础模型。经过互联网规模视频、真实驾驶和机器人数据以及 3D 模拟的预训练。
**English / Original**
But the challenge is clear. The physical world is diverse and unpredictable. Collecting real-world training data is slow and costly, and it's never enough. The answer is synthetic data. It starts with NVIDIA Cosmos, an open frontier world foundation model for physical AI. Pre-trained on internet-scale video, real driving and robotics data, and 3D simulation.
## 00:32:04
**中文**
Cosmos 学会了对世界的统一表示，能够协调语言、图像、3D 和动作。它执行物理 AI技能，如生成、推理和轨迹预测。 Cosmos 从单个图像生成逼真的视频。来自 3D 场景描述、物理连贯的运动。从驾驶遥测和传感器日志到环绕视频。
**English / Original**
Cosmos learned a unified representation of the world, able to align language, images, 3D, and action. It performs physical AI skills like generation, reasoning, and trajectory prediction. From a single image, Cosmos generates realistic video. From 3D scene descriptions, physically coherent motion. From driving telemetry and sensor logs, surround video.
## 00:32:39
**中文**
从规划模拟器、多摄像机环境。或者从场景提示来看，它使边缘案例变得栩栩如生。开发人员可以在 Cosmos 中运行交互式闭环模拟。当采取行动时，世界就会做出反应。Cosmos 会进行推理。它分析边缘场景，将其分解为熟悉的物理交互，并解释接下来可能发生的情况。
**English / Original**
From planning simulators, multi-camera environments. Or from scenario prompts, it brings edge cases to life. Developers can run interactive closed-loop simulations in Cosmos. When actions are made, the world responds. Cosmos reasons. It analyzes edge scenarios, breaks them down into familiar physical interactions, and reasons about what could happen next.
## 00:33:13
**中文**
Cosmos 将计算转化为数据，训练长尾自动驾驶汽车，以及机器人如何适应各种场景。我知道，这太不可思议了。 Cosmos是全球领先的基础模型，世界基础模型。它已被下载数百万次，在世界各地使用。让世界为物理 AI的新时代做好准备。我们自己也用它。
**English / Original**
Cosmos turns compute into data, training AVs for the long tail, and robots how to adapt for every scenario. I know, it's incredible. Cosmos is the world's leading foundation model, world foundation model. It's been downloaded millions of times, used all over the world. Getting world getting the world ready for this new era of physical AI. We use it ourselves as well.
## 00:33:49
**中文**
我们自己用它来制造我们的自动驾驶汽车。使用它来生成场景并进行评估，我们可以拥有一些东西，可以让我们有效地行驶数十亿、数万亿英里，但在计算机内进行。我们已经取得了巨大的进步。今天，我们宣布推出 Alpamayo，这是世界上第一个会思考、推理的自动驾驶汽车人工智能。
**English / Original**
We use it ourselves to create our self-driving car. Using it for scenario generation and using it for evaluation, we could have something that allows us to effectively travel billions, trillions of miles, but doing it inside a computer. And we've made enormous progress. Today, we're announcing Alpha Mylo, the world's first thinking, reasoning autonomous vehicle AI.
## 00:34:23
**中文**
Alpamayo 进行了端到端的训练，从相机输入到驱动输出。摄像头通过人类演示自行驾驶了很多很多英里，而我们有很多很多由 Cosmos 生成的里程。除此之外，数十万个示例都被非常非常仔细地标记，以便我们可以教汽车如何驾驶。 Alpamayo 做了一些非常特别的事情。
**English / Original**
Alpha Mylo is trained end to end, literally from camera in to actuation out. The camera in, lots and lots of miles that are driven by itself using human demonstration, and we have lots and lots of miles that are generated by Cosmos. In addition to that, hundreds of thousands of examples are labeled very, very carefully so that we could teach the car how to drive. Alpha Mylo does something that's really special.
## 00:34:53
**中文**
它不仅接受传感器输入并激活方向盘、刹车和加速，还推理将要采取的行动。它告诉你它将采取什么行动，采取该行动的原因，当然还有轨迹。所有这些都是直接耦合的，并通过大量人类训练数据和 Cosmos 生成数据的组合进行非常具体的训练。
**English / Original**
Not only does it take sensor input and activates steering wheel, brakes, and and acceleration, it also reasons about what action it's about to take. It tells you what action it's going to take, the reasons by which it came about that action, and then of course the trajectory. All of these are coupled directly and trained very specifically by a large combination of human trained and as well as Cosmos generated data.
## 00:35:27
**中文**
其结果真是令人难以置信。你的汽车不仅会按照你期望的方式行驶，而且它会如此自然地行驶，因为它直接从人类演示者那里学习，而且在每一个场景中，当它遇到场景时，它会推理它，告诉你它要做什么，并推理它要做什么。
**English / Original**
The result of it is just really incredible. Not only does your car drive as you would expect it to drive, and it drives so naturally because it learned directly from human demonstrators, but in every single scenario when it comes up to the scenario, it reasons about it, tells you what it's going to do, and it reasons about what it's what it's about to do.
## 00:35:46
**中文**
现在，这一点之所以如此重要，是因为驾驶的长尾效应。我们不可能简单地收集每个国家可能发生的所有情况以及所有人口可能发生的每种情况的所有可能情况。然而，如果将每个场景分解为一大堆其他较小的场景，那么很可能每个场景都对您来说是很正常的理解。因此，这些长尾巴将被分解为汽车知道如何处理的非常正常的情况。它只需要推理一下即可。那么，让我们来看看。
**English / Original**
Now, the reason why this is so important is because of the long tail of driving. There it's impossible for us to simply collect every single possible scenario for everything that could ever happen in every single country and every single circumstance that's possibly ever going to happen for all of population. However, it is very unlikely it's very likely that every scenario, if decomposed into a whole bunch of other smaller scenarios, are quite normal for you to understand. And so, these long tails will be decomposed into quite normal circumstances that the car knows how to deal with. It just needs to reason about it. And so, let's take a look.
## 00:36:28
**中文**
你将要看到的一切都只是一瞬间。全程无需手动操作。前往目的地的路线。系好安全带。
**English / Original**
Everything you're about to see is one shot. It's a no hands. Routing to your destination. Buckle up.
## 00:39:12
**中文**
你已经到了。我们八年前开始研究自动驾驶汽车，原因是我们很早就推断深度学习和人工智能将重塑整个计算。如果我们要了解如何导航自己以及如何引导行业走向这个新的未来，我们就必须善于构建整个堆栈。
**English / Original**
You have arrived. We started working on self-driving cars 8 years ago, and the reason for that is because we reasoned early on that deep learning and artificial intelligence was going to reinvent the entire computing And if we were ever going to understand how to navigate ourselves and how to guide the industry towards this new future, we have to get good at building the entire stack.
## 00:39:51
**中文**
嗯，正如我之前提到的，人工智能是一个五层蛋糕。最底层是土地、电力与建筑。就机器人技术而言，最底层是汽车。它上面的下一层是芯片。 GPU、网络芯片、CPU，所有这些东西。上面的下一层是基础设施。在这个特殊情况下，正如我在物理 AI中提到的那样，该基础设施就是 Omniverse 和 Cosmos。
**English / Original**
Well, as I mentioned earlier, AI is a five-layered cake. The lowest layer is land, power, and shell. In the case of robotics, the lowest layer is the car. The next layer above it is chips. GPUs, networking chips, CPUs, all that kind of stuff. The next layer above that is the infrastructure. That infrastructure, in this particular case, as I mentioned with physical AI, is Omniverse and Cosmos.
## 00:40:18
**中文**
然后上面是模型。就我刚才向您展示的上述模型而言，这里的模型称为 Alpamayo。今天的 Alpamayo 是开源的。我们这项令人难以置信的工作，花费了数千人的努力。我们的 AV 团队有几千人。客观地说，我们的搭档呃奥拉，我认为奥拉就在观众席的某个地方。
**English / Original**
And then above that are the models. And in the case of the models above that I just shown you, the model here is called Alpha Miao. And Alpha Miao today is open-sourced. We This incredible body of work, it took several thousand people. Our AV team is several thousand people. Just to put in perspective, our partner uh Ola, I think Ola is here in the audience somewhere.
## 00:40:48
**中文**
嗯，梅赛德斯五年前同意与我们合作，让这一切成为可能。我们想象有一天道路上的 10 亿辆汽车将全部实现自动驾驶。你可以让它成为一辆机器人出租车，你正在策划并从某人那里租用，或者你可以拥有它，它会为你自动驾驶，或者你可以决定自己驾驶。
**English / Original**
Uh Mercedes agreed to partner with us 5 years ago to go make all of this possible. We imagine that someday a billion cars on the road will all be autonomous. You could either have it be a robo-taxi that you're you're you're orchestrating and and renting from somebody, or you could own it and it's driving for you driving by itself, or you could decide to drive for yourself.
## 00:41:10
**中文**
因此，每辆汽车都将具有自动驾驶能力。每辆汽车都将由人工智能驱动。所以，本例中的模型层是 Alpamayo，上面的应用程序是 Mercedes-Benz。好的？所以，这个整个堆栈是我们第一个 NVIDIA 的第一个整个堆栈的努力。我们一直在努力，我们很高兴 NVIDIA 的第一款 AV 汽车将在第一季度上路，然后在第二季度进入欧洲，在美国，然后在第二季度进入欧洲，我认为第三季度和第四季度将进入亚洲。最重要的是，我们将继续更新它的下一个版本
**English / Original**
And so, but every single car will have autonomous vehicle capability. Every single car will be AI-powered. And so, the the the model layer in this case is Alpha Miao, and the application above that is the Mercedes-Benz. Okay? And so, so this entire stack is our first NVIDIA first entire stack endeavor. And we've been working on it for this entire time, and we're so happy that the first AV car from NVIDIA is going to be on the road in Q1, and then it goes Europe in Q2 here in the United States in Q1, then Europe in Q2, and I think it's Asia in Q3 and Q4. And the powerful thing is that we're going to keep on updating it with next next versions of
## 00:41:53
**中文**
Alpamayo 以及之后的版本。
**English / Original**
Alpha Miao and versions after that.
## 00:41:55
**中文**
现在我心中毫无疑问这将成为最大的机器人行业之一，我很高兴我们为此而努力。它教会了我们如何帮助世界其他地方构建机器人系统的大量知识。深刻理解并知道如何自己构建它，自己构建整个基础设施，并知道机器人系统需要什么样的芯片。在这个特殊的情况下，双Orion、下一代双Thors这些处理器是专为机器人系统设计的，并且是为安全最高级别的安全能力而设计的。这辆车刚刚获得评级，刚刚投入生产。
**English / Original**
There's no question in my mind now that this is going to be one of the largest robotics industries, and I'm so happy that we worked on it. And it taught us enormous amount about how to help the rest of the world build robotic systems. That deep understanding and knowing how to build it ourselves, building the entire infrastructure ourselves, and knowing what kind of chips a robotic system would would need. In this particular particular case, dual Orions, the next generation dual Thors, these processors are designed for robotic systems, and was designed for the safe highest level of safety capability. This car just got rated it just went to production.
## 00:42:38
**中文**
梅赛德斯-奔驰 CLA 刚刚被 NCAP 评为全球最安全的汽车。据我所知，这是唯一一个每行代码、芯片、系统、每一行代码都经过安全认证的系统。整个模型系统基于传感器多样化且冗余，自动驾驶汽车堆栈也是如此。 Alpamayo 堆栈经过端到端训练，拥有令人难以置信的技能。
**English / Original**
The Mercedes-Benz CLA was just rated by NCAP the world's safest car. It is the only system that I know that has every single line of code, the chip, the system, every line of code safety certified. The entire model system is based on a sensors are diverse and redundant, and so is the self-driving car stack. The Alpha Miao stack is trained end to end and has incredible skills.
## 00:43:13
**中文**
然而，直到你永远驾驶它之前，没有人知道它会是绝对安全的。因此，我们采用另一个软件堆栈（底层是整个 AV 堆栈）来保护这一点。整个 AV 堆栈的构建是完全可追溯的，我们花了大约 5 年的时间来构建它，实际上大约花了 6 7 年的时间来构建第二个堆栈。
**English / Original**
However, nobody knows until you drive it forever that it's going to be perfectly safe. And so, the we the way we guardrail that is with another software stack, an entire AV stack underneath. That entire AV stack is built to be fully traceable, and it's taken us some 5 years to build that, some 6 7 years actually, to build that second stack.
## 00:43:35
**中文**
这两个软件堆栈相互镜像，然后我们有一个政策和安全评估决定，这是我非常有信心并且可以非常安全驾驶的原因吗？如果是这样，我会让阿尔法缪来做。如果我对这种情况不太有信心，并且安全嗯政策评估者决定我们将回到一个非常简单、更安全的护栏系统，那么它就会回到经典的 AV 堆栈。
**English / Original**
These two software stacks are mirroring each other, and then we have a policy and safety evaluated decide, is this something that I'm very confident and can reason about driving very safely? If so, I'm going to have Alpha Miao do it. If it's a circumstance that I'm not very confident in and the safety um policy evaluator decide that we're going to go back to what's a very a simpler, safer guardrail system, then it goes back to the classical AV stack.
## 00:44:02
**中文**
我们是世界上唯一同时运行这两个自动驾驶堆栈的汽车，所有安全系统都应该具有多样性和冗余性。嗯，我们的愿景是有一天每辆汽车、每辆卡车都将实现自动驾驶，我们一直在为这个未来而努力。当然，就梅赛德斯-奔驰而言，整个堆栈是垂直整合的。
**English / Original**
We're the only car in the world with both of these AV stacks running, and all safety systems should have diversity and redundancy. Well, our vision is that someday every single car, every single truck will be autonomous, and we've been working towards that future. This entire stack is vertically integrated, of course, in the case of Mercedes-Benz.
## 00:44:20
**中文**
我们一起构建了整个堆栈。我们要部署汽车。我们要操作堆栈。只要我们还活着，我们就会维护这个堆栈。然而，就像我们作为一家公司所做的其他事情一样，我们构建了整个堆栈，但整个堆栈对生态系统开放。与我们合作构建 L4 和自动驾驶出租车的生态系统正在不断扩大，并且遍及各地。
**English / Original**
We built the entire stack together. We're going to deploy the car. We're going to operate the stack. We're going to maintain the stack for as long as we shall live. However, like everything else we do as a company, we build the entire stack, but the entire stack is open for the ecosystem. And these the ecosystem working with us to build L4 and robo-taxis is expanding and it's going everywhere.
## 00:44:44
**中文**
我完全希望这会是，这对我们来说已经是一项巨大的业务了。这对我们来说是一项巨大的业务，因为他们用它来训练我们的训练数据、处理数据和训练他们的模型。他们用它来生成合成数据。在某些情况下，在某些公司的某些汽车中，他们几乎只是制造计算机，即汽车内部的芯片。有些公司与我们进行全栈合作。有些公司与我们合作其中的一部分。好的？因此，您决定使用多少并不重要。你知道，我唯一的要求是你尽可能使用一点 NVIDIA 技术。整个事情都是开放的。
**English / Original**
I fully expect this to be Well, this is already a giant business for us. It's a giant business for us because they use it for training our training data, processing data, and training their models. They use it for synthetic data generation. In some cases, in some car in some companies, they build pretty much just build the computers, the chips that are inside the car. And some companies work with us full stack. Some companies work with us some partial part of that. Okay? So, it doesn't matter how much you decide to use. You know, my only request is you use a little bit of video wherever you can. And the entire thing is open.
## 00:45:19
**中文**
现在，这将是第一个大规模的主流嗯物理 AI市场。现在，我想我们都可以完全同意。从非自动驾驶汽车到自动驾驶汽车的拐点可能就在这个时候发生。在接下来的十年里，我相当肯定世界上非常非常大比例的汽车将是自动驾驶或高度自动驾驶的。
**English / Original**
Now, this is going to be the first large-scale mainstream um AI physical AI market. And this is now, I think we can all agree, fully here. And this inflection point of going from not autonomous vehicles to autonomous vehicles is probably happening right about this time. In in the next 10 years, I'm fairly certain a very, very large percentage of the world's cars will be autonomous or highly autonomous.
## 00:45:48
**中文**
但这是我刚刚在使用三台计算机时描述的基本技术，使用合成数据生成和模拟，适用于每种形式的机器人系统。它可能是一个只是一个关节器、一个操纵器的机器人。也许它是一个移动机器人。也许它是一个完全人形机器人。因此，机器人系统的下一个时代的下一个旅程将是机器人。这些机器人将有各种不同的尺寸。我还邀请了一些朋友。
**English / Original**
But this is this basic technique that I just described in using the three computers, using the synthetic data generation and simulation, applies to every form of robotic systems. It could be a robot that is just an articulator, a manipulator. Maybe it's a mobile robot. Maybe it's a fully humanoid robot. And so, the next journey the next era for robotic systems is going to be, you know, robots. And these robots are going to come in all kinds of different sizes. And And I invited some friends.
## 00:46:21
**中文**
嘿伙计们。赶快。你告诉 R2-D2 你要来这里了吗？你有吗？还有C-3PO？好的。好吧，到这里来。在此之前，其中之一就是真正的Jetson。他们里面有小型 Jetson 电脑。他们在 Omniverse 内接受训练。那这个又如何呢？让我们向大家展示你们学习如何成为机器人的模拟器。你们想要看看吗？
**English / Original**
Hey guys. Hurry up. Did you tell R2-D2 you're going to be here? Did you? And C-3PO? Okay. All right, come here. Before Now, one of the things that one of the things that's really You have Jetsons. They have little Jetson computers inside them. They're trained inside Omniverse. And how about this? Let's show everybody the simulator that you were that you guys learn how to how to be robots in. You want You guys want to look at that?
## 00:46:54
**中文**
好吧，让我们看看。请运行它。
**English / Original**
Okay, let's look at that. Run it, please.
## 00:47:57
**中文**
这不是很神奇吗？这就是你学习成为机器人的方式。这一切都是你在 Omniverse 内完成的。机器人模拟器名为 Isaac。Isaac Sim 和 Isaac Lab。任何想要建造机器人的人，你知道，没有人能做到，没有人会像你一样可爱。但现在我们都看看所有这些看看我们建造机器人的所有这些朋友。我们正在建造大型的。不，就像我说的，没有人像你们那么可爱，但我们有 Neurobot，我们那里有 Agibot。
**English / Original**
Isn't that amazing? That's how you learn to be a robot. You did it all inside Omniverse. And the robot simulator is called Isaac. Isaac Sim and Isaac Lab. And anybody who wants to build a robot, you know, nobody could Nobody's going to be as cute as you. But now we have all Look at all these Look at all these friends that we have building robots. We have We're building big ones. No Like I said, nobody's as cute as you guys are, but we have Neurobot, and we have Agibot over there.
## 00:48:30
**中文**
你知道，我们这里有 LG。他们刚刚宣布了一款新机器人。卡特彼勒，他们拥有有史以来最大的机器人。那个人送食物到你家。这与 Uber Eats 优食有关。还有那个冲浪机器人，我喜欢那些家伙。Agility、Boston Dynamics，令人难以置信。你有手术机器人。你有Franka的机械手机器人。
**English / Original**
You know, we have LG over here. They just announced a new robot. Caterpillar, they've got the largest robots ever. That one delivers food to your house. That's connected to Uber Eats. And that Surf robot, I love those guys. Agility, Boston Dynamics, incredible. You got surgical robots. You got manipulator robots from Franka.
## 00:48:55
**中文**
你有Universal Robots的机器人，数量惊人的不同机器人。所以，这是下一章。未来我们将更多地讨论机器人技术，但最终不仅仅是机器人。我知道一切都与你们有关。这是关于到达那里。物理 AI和AI 物理将带来世界上最重要的行业之一，而这个行业正是我们所有人开创的行业。
**English / Original**
You got Universal Robotics robot, incredible number of different robots. And so, this is the next chapter. We're going to talk a lot more about robotics in the future, but it's not just about the robots in the end. I know everything's about you guys. It's about getting there. And one of the one of the most important industries in the world that will be revolutionized by physical AI and AI physics is the industry that started all of us.
## 00:49:22
**中文**
在 NVIDIA，如果没有我即将交谈的公司，这一切都是不可能的。我很高兴他们所有人都从 Cadence 开始，将加速 Cadence CUDA-X 集成到其所有模拟和求解器中的一切。他们拥有 NVIDIA 物理 AI，将用于不同的实体工厂及工厂模拟。
**English / Original**
At NVIDIA, it wouldn't be possible if not for the companies that I'm about to talk to. And I'm so happy that all of them, starting with Cadence, is going to accelerate everything Cadence CUDA X integrated into all of their simulations and solvers. They've got NVIDIA physical physical AIs that they're going to use for for different physical plants and plant simulations.
## 00:49:45
**中文**
你可以将AI 物理集成到这些系统中。因此，无论是 EDA 还是 SDA，以及未来的机器人系统，我们将拥有基本相同的技术，使你们成为可能。现在彻底改变这些设计堆栈。 Synopsys，没有Synopsys 要知道，Synopsys和Cadence在芯片设计领域是完全不可缺少的。 Synopsys 在逻辑设计和 IP 领域处于领先地位。呃，就 Cadence 而言，他们领导物理设计、布局和布线、呃以及仿真和验证。 Cadence 在仿真和验证方面的表现令人难以置信。两人都在动
**English / Original**
You got AI physics being integrated into these systems. So, whether it's an EDA or SDA and in the future robotic systems, we're going to have basically the same technology that made you guys possible. Now completely revolutionize these design stacks. Synopsys, without Synopsys You know, Synopsys and Cadence are completely completely indispensable in the world of chip design. Uh Synopsys is uh leads in uh in logic design and and IP. Uh in the case of Cadence, they lead physical design, the place and route, uh and emulation and verification. Cadence is incredible at emulation and verification. Both of them are moving
## 00:50:29
**中文**
进入系统设计和系统仿真的世界。因此，将来我们将在 Cadence 和 Synopsys 内部设计您的芯片。我们将设计您的系统并模拟整个系统并模拟这些工具内的所有内容。那就是你的未来。我们将给予，是的，你将在这些平台内出生。非常神奇，对吧？因此，我们很高兴能够与这些行业合作。正如我们将 NVIDIA 集成到 Palantir 和 ServiceNow 中一样，我们也将 NVIDIA 集成到计算最密集的模拟行业 Synopsys 和 Cadence 中。今天我们宣布
**English / Original**
into the world of system design and system simulation. And so, in the future we're going to design your chips inside Cadence and inside Synopsys. We're going to design your systems and emulate the whole thing and simulate everything inside these tools. That's your future. We're going to give Yeah, you're going to be born inside these inside these platforms. Pretty amazing, right? And so, we're so happy that we're working with these these industries. Just as we've integrated NVIDIA into Palantir and ServiceNow, we're integrating NVIDIA into the most computationally intensive simulation industries, Synopsys and Cadence. And today we're announcing that
## 00:51:10
**中文**
西门子也在做同样的事情。
**English / Original**
Siemens is also doing the same thing.
## 00:51:12
**中文**
我们要把CUDA-X物理AI、智能体 AI、NeMo、NeMotron深度融入西门子的世界。原因是这样的。首先，我们设计芯片，未来所有这些都会由NVIDIA加速。你会为此感到非常高兴的。我们将有智能体芯片设计师和系统设计师与我们合作，帮助我们进行设计，就像我们今天有智能体软件工程师帮助我们的软件工程师编码一样。因此，我们将拥有智能体芯片设计师和系统设计师。我们将在其中创造你。
**English / Original**
We're going to integrate CUDA X physical AI, agentic AI, Nemo, Nemotron deeply integrated into the world of Siemens. And the reason for that is this. First, we design the chips and all of it in the future will be accelerated NVIDIA. You're going to be very happy about that. We're going to have agentic chip designers and system designers working with us, helping us do design just as we have agentic software engineers helping our software engineers code today. And so, we'll have agentic chip designers and system designers. We're going to create you inside this.
## 00:51:49
**中文**
但接下来，我们必须打造你。我们必须建造工厂，为您制造产品的工厂。我们必须设计组装你们所有人的生产线。这些制造工厂本质上将是巨大的机器人。难以置信，不是吗？我知道。我知道。所以，你将在计算机中被设计。
**English / Original**
But then, we have to build you. We have to build the plants, the factories that make manufacture you. We have to design the manufacturing lines that assemble all of you. And these manufacturing plants are going to be essentially gigantic robots. Incredible, isn't that right? I know. I know. And so, you're going to be designed in a computer.
## 00:52:15
**中文**
你将在电脑中被创造出来。在你不得不花时间处理重力之前，你很快就会在计算机上接受测试和评估。我知道。你知道如何应对重力吗？你能跳吗？你能跳吗？好的。好的。不要炫耀。好的。所以，所以这个所以，现在这个行业使 NVIDIA 成为可能，我很高兴现在
**English / Original**
You're going to be made in a computer. You're going to be tested and evaluated in a computer long before long before you have to spend any time dealing with gravity. I know. Do you know how to deal with gravity? Can you jump? Can you jump? Okay. All right. Don't show off. Okay. So, so this So, now the industry the industry that made NVIDIA possible where I'm just so happy that the now the
## 00:52:58
**中文**
我们正在创造的技术已经达到了一定的复杂程度和能力，我们现在可以帮助他们彻底改变他们的行业。因此，从他们开始的事情开始，我们现在有机会回去帮助他们彻底改变自己的事情。让我们看一下我们将与西门子一起做的事情。
**English / Original**
technology that we're creating is at a level of sophistication and capability that we can now help them revolutionize their industry. And so, what started with with with them, we now have the opportunity to go back and and help them revolutionize theirs. Let's take a look at the stuff that we're going to do with Siemens.
## 00:53:16
**中文**
快点。物理 AI的突破正在让人工智能从屏幕转移到我们的物理世界。恰逢世界各地建造各种芯片、计算机、救生药物和人工智能工厂。随着全球劳动力短缺的加剧，我们比以往任何时候都更需要由物理 AI和机器人技术驱动的自动化。人工智能与世界上最大的实体行业的结合，是 NVIDIA 和西门子合作伙伴关系的基础。
**English / Original**
Come on. Breakthroughs in physical AI are letting AI move from screens to our physical world. And just in time as the world builds factories of every kind for chips, computers, life-saving drugs, and AI. As the global labor shortage worsens, we need automation powered by physical AI and robotics more than ever. This, where AI meets the world's largest physical industries, is the foundation of NVIDIA and Siemens partnership.
## 00:53:53
**中文**
近两个世纪以来，西门子建设了世界工业。现在，它正在为人工智能时代重塑它。西门子正在将 NVIDIA CUDA-X 库 AI 模型和 Omniverse 集成到其 EDA CAE 和数字孪生工具和平台产品组合中。我们共同将物理 AI引入整个工业生命周期。
**English / Original**
For nearly two centuries, Siemens has built the world's industries. And now, it is reinventing it for the age of AI. Siemens is integrating NVIDIA CUDA X libraries AI models and Omniverse into its portfolio of EDA CAE and digital twin tools and platforms. Together, we're bringing physical AI to the full industrial life cycle.
## 00:54:27
**中文**
从设计和模拟到生产和运营。我们正处于新工业革命的开端。 NVIDIA 和西门子为下一个工业时代打造的物理 AI 时代。难以置信，对吧，伙计们？你怎么认为？好的。我会紧紧抓住。紧紧抓住。所以，这就是你知道的，如果你看看世界上的模型，毫无疑问 OpenAI 是当今领先的token生成器。
**English / Original**
From design and simulation to production and operations. We stand at the beginning of a new industrial revolution. The age of physical AI built by NVIDIA and Siemens for the next age of industries. Incredible, right, guys? What do you think? All right. I'll hang on tight. Just hang on tight. And so, so this is you know, if you look at look at the world's models there's no question OpenAI is the the leading token generator today.
## 00:55:08
**中文**
生成的 OpenAI token比其他任何token都多。第二大组第二大可能是开放模型。我的猜测是，随着时间的推移，由于有如此多的公司、如此多的研究人员、如此多不同类型的领域和模式，开源模型将成为迄今为止最大的。
**English / Original**
More to more OpenAI tokens are generated than just about anything else. The second largest group the second largest probably open models. And my guess is that over time, because there are so many companies, so many researchers, so many different types of domains and modalities that open source models will be by far the largest.
## 00:55:27
**中文**
让我们来谈谈一个非常特别的人。你们想这么做吗？我们来谈谈Vera Rubin。Vera Rubin 是的，继续吧。她是一位美国天文学家。她是第一个观察到的人，她注意到星系尾部的移动速度与星系中心的移动速度一样快。好吧，我知道这没有意义。这毫无意义。牛顿物理学会说，就像太阳系一样，离太阳较远的行星绕太阳运行的速度比离太阳较近的行星慢。
**English / Original**
Let's talk about somebody really special. You guys want to do that? Let's talk about Vera Rubin. Vera Rubin Yeah, go ahead. She's an American astronomer. She was the first to observe She noticed that the tails of the galaxies were moving about as fast as the center of the galaxies. Well, I know it makes no sense. It makes no sense. Newtonian physics would say, just like the solar system, the planets further away from the sun is circling the sun slower than the planets closer to the sun.
## 00:56:04
**中文**
因此，除非存在看不见的物体，否则这种情况的发生是没有意义的，我们称它们为她发现的暗体暗物质，嗯，尽管我们看不到它，但它占据了空间。因此，我们以Vera Rubin 的名字命名我们的下一台计算机。这不是个好主意吗？我知道。好的。 Vera Rubin 旨在解决我们面临的这一基本挑战。
**English / Original**
And therefore, it makes no sense that this happens unless there's invisible bodies, we call them she discovered dark body dark matter, um that occupies space even though we don't see it. And so, Vera Rubin is the person that we named our next computer after. Isn't that a good idea? I know. Okay. Vera Rubin is designed to address this fundamental challenge that we have.
## 00:56:35
**中文**
人工智能所需的计算量正在猛增。对 NVIDIA GPU 的需求正在飙升。它之所以飞速增长，是因为模型每年都以 10 倍的数量级增长。更不用说，正如我提到的，o1 的推出是人工智能的转折点。推理现在不再是一次性答案，而是一个思考过程。为了教会人工智能如何思考，在训练后引入了强化学习和非常重要的计算。不久之后，它不再是监督微调或以其他方式称为
**English / Original**
The amount of computation necessary for AI is skyrocketing. The demand for NVIDIA GPUs is skyrocketing. It's skyrocketing because models are increasing by a factor of 10 an order of a magnitude every single year. And not to mention, as I mentioned, 01's introduction was an inflection point for AI. Instead of a one-shot answer, inference is now a thinking process. And in order to teach the AI how to think, reinforcement learning and very significant computation was introduced into post-training. It wasn't no long It's no longer supervised fine-tuning or otherwise known as
## 00:57:17
**中文**
模仿学习或监督训练。现在你有了强化学习，本质上是计算机试验本身尝试不同的迭代，学习如何执行任务。因此，预训练、训练后、测试时扩展的计算量呈爆炸式增长。现在，我们所做的每一个推论，而不只是一次，你都可以看到人工智能思考的token数量，我们对此表示赞赏。它思考的时间越长，通常会产生更好的答案。因此，测试时间的扩展导致生成的token数量每年增加 5 倍。更不用说与此同时，人工智能的竞赛已经开始。
**English / Original**
imitation learning or supervision training. You now have reinforcement learning, essentially the computer trial trying different iterations itself, learning how to perform a task. The amount of computation for pre pre-training, for post-training, for test time scaling has exploded as a result of that. And now every single inference that we do, instead of just one shot, the number of tokens you could just see the AIs think, which we appreciate. The longer it thinks, often times it produces a better answer. And so, test time scaling causes the number of tokens to be generated increase by 5x every single year. Not to mention meanwhile, the race is on for AI.
## 00:58:01
**中文**
每个人都在努力达到新的水平。每个人都在努力进入下一个前沿。每当他们到达上一代 AI token的下一个前沿时，成本就会开始每年下降约 10 倍。每年10倍的下降实际上告诉你一些不同的事情。据说比赛非常激烈。每个人都在努力进入下一个水平，而有人正在进入下一个水平。因此，所有这一切都是一个计算问题。计算速度越快，就能越早达到下一个前沿的新水平。
**English / Original**
Everybody's trying to get to the next level. Everybody's trying to get to the next frontier. And every time they get to the next frontier the last generation AI tokens, the cost starts to starts to decline about a factor of 10x every year. The 10x decline every year is actually telling you something different. It's saying that the race is so intense. Everybody's trying to get to the next level and somebody is getting to the next level. And so, therefore all of it is a computing problem. The faster you compute, the sooner you can get to the next level of the next frontier.
## 00:58:34
**中文**
所有这些事情都是同时发生的。因此，我们决定每年都必须提高计算技术的水平。没有一年落后。现在，我们已经在一年半前推出 GB200。目前，我们正在全面生产GB300。如果Vera Rubin（Vera Rubin）今年能及时上市，它现在就必须投入生产。所以，今天我可以告诉你，Vera Rubin 正在全面制作中。
**English / Original**
All of these things are simultaneously happening at the same time. And so, we decided that we have to advance the state of the art of computation every single year. Not one year left behind. And now, we've been shipping GB200s year and a half ago. Right now, we're in full-scale manufacturing of GB300. And if Vera Rubin is going to be in time for this year, it must be in production by now. And so, today I can tell you that Vera Rubin is in full production.
## 00:59:15
**中文**
你们想看看Vera Rubin吗？好的。快点。请播放一下。Vera Rubin 及时赶到，迎接人工智能的下一个前沿领域。这是我们如何构建它的故事。该架构是一个由六颗芯片组成的系统，经过精心设计，可以作为一个整体工作，诞生于极端的协同设计。它从 Vera 开始，这是一款定制设计的 CPU，其性能是上一代的两倍。
**English / Original**
You guys want to take a look at Vera Rubin? All right. Come on. Play it, please. Vera Rubin arrives just in time for the next frontier of AI. This This the story of how we built it. The architecture, a system of six chips engineered to work as one, born from extreme co-design. It begins with Vera, a custom-designed CPU, double the performance of the previous generation.
## 00:59:45
**中文**
还有 Rubin GPU。 Vera 和 Rubin 从一开始就是共同设计的，可以更快、更低延迟地双向、一致地共享数据。然后，17,000 个组件聚集在 Vera Rubin 计算板上。在 Vera CPU 和两个 Rubin GPU 完成组装之前，高速机器人以微米精度放置组件。
**English / Original**
And the Rubin GPU. Vera and Rubin are co-designed from the start to bidirectionally and coherently share data faster and with lower latency. Then, 17,000 components come together on a Vera Rubin compute board. High-speed robots place components with micron precision before the Vera CPU and two Rubin GPUs complete the assembly.
## 01:00:14
**中文**
能够提供 100 petaflops 的 AI，是其前身的五倍。人工智能需要快速数据。 ConnectX-9 为每个 GPU 提供每秒 1.6 太比特的横向扩展带宽。 BlueField-4 DPU 减轻了存储和安全性的负担，因此计算完全专注于人工智能。 Vera Rubin 计算托盘经过完全重新设计，没有电缆、软管或风扇。
**English / Original**
Capable of delivering 100 petaflops of AI, five times that of its predecessor. AI needs data fast. ConnectX-9 delivers 1.6 terabits per second of scale-out bandwidth to each GPU. BlueField-4 DPU offloads storage and security, so compute stays fully focused on AI. The Vera Rubin compute tray, completely redesigned with no cables, hoses, or fans.
## 01:00:48
**中文**
配备 BlueField-4 DPU、八个 ConnectX-9 NIC、两个 Vera CPU 和四个 Rubin GPU。 Vera Rubin AI 超级计算机的计算构建块。接下来是第六代NVLink交换机。移动的数据量超过全球互联网，连接 18 个计算节点，最多可扩展至 72 个 Rubin GPU 作为一个运行。然后是 Spectrum-X 以太网光子学。
**English / Original**
Featuring a BlueField-4 DPU, eight ConnectX-9 NICs, two Vera CPUs, and four Rubin GPUs. The compute building block of the Vera Rubin AI supercomputer. Next, the sixth-generation NVLink switch. Moving more data than the global internet, connecting 18 compute nodes, scaling up to 72 Rubin GPUs operating as one. Then, Spectrum-X Ethernet photonics.
## 01:01:24
**中文**
全球首款具有 512 个通道和 200 Gb 能力的集成光学器件的以太网交换机，可将数千个机架扩展到人工智能工厂。自设计开始以来，经过 15,000 个工程师年的努力，第一台 Vera Rubin NVL72 机架上线。六颗突破性芯片、18 个计算托盘、9 个 NVLink 交换机托盘、220 万亿个晶体管，重近两吨。
**English / Original**
The world's first Ethernet switch with 512 lanes and 200 gigabit capable co- packaged optics scale out thousands of racks into an AI factory. 15,000 engineer years since design began, the first Vera Rubin MVL-72 rack comes online. Six breakthrough chips, 18 compute trays, nine NVLink switch trays, 220 trillion transistors, weighing nearly two tons.
## 01:01:58
**中文**
迈向人工智能下一个前沿的巨大飞跃。Rubin来了。你们觉得怎么样？这是Rubin pod。 16 个机架中有 1,152 个 GPU。如您所知，每个架子上有 72 个 Rubin GPU。每个 Rubin GPU 都是两个连接在一起的实际 GPU 芯片。我要把它展示给你看。但是，有几件事我稍后会告诉你。
**English / Original**
One giant leap to the next frontier of AI. Rubin is here. What do you guys think? This is a Rubin pod. 1,152 GPUs in 16 racks. Each one of the racks, as you know, has uh 72 Rubins. Each one of the Rubins is two actual GPU dies connected together. I'm going to show it to you. But, there are several things that, well, I'll tell you later.
## 01:02:42
**中文**
嗯，我们设计了六种不同的芯片。首先，我们公司内部有一个规定，而且是一个很好的规定。新一代芯片的变化不应超过一两个。但是，问题是这样的。正如您所看到的，我们描述了所描述的每一个芯片中的晶体管总数，并且我们知道摩尔定律已经大大放缓。因此，我们年复一年获得的晶体管数量不可能跟上10倍大的模型。它不可能跟上每年五倍的token生成速度。它不可能跟上token成本下降的事实
**English / Original**
Well, we designed six different chips. First of all, we have a rule inside our company, and it's a good rule. No new generation should have more than one or two chips change. But, the problem is this. As you could see, we were describing the total number of transistors in each one of the chips that were being described, and we know that Moore's law has largely slowed. And so, the number of transistors we can get year after year after year can't possibly keep up with the 10 times larger models. It can't possibly keep up with five times per year more tokens generated. It can't possibly keep up with the fact that cost decline of the tokens are going to
## 01:03:24
**中文**
如此咄咄逼人。
**English / Original**
be so aggressive.
## 01:03:26
**中文**
除非我们部署积极、极端的协同设计，基本上在所有芯片、整个堆栈上同时进行创新，否则行业不可能跟上这种速度继续前进。这就是为什么我们决定在这一代，我们别无选择，只能重新设计每个芯片。现在，我们刚才描述的每一个芯片本身都可以成为一场新闻发布会，并且有一家公司的产品致力于在过去做到这一点。
**English / Original**
It is impossible to keep up with those kind of rates for the industry to continue to advance unless we deployed aggressive, extreme co-design, basically innovating across all of the chips, across the entire stack, all at the same time. Which is the reason why we decided that this generation, we had no choice but to design every chip over again. Now, every single chip that we were describing just now can be a press conference in all in itself, and there's an entire company whose product dedicated to doing that back in the old days.
## 01:03:59
**中文**
其中每一项都是完全革命性的，并且是同类产品中最好的。 Vera CPU，我为它感到非常自豪。在功耗受限的环境中，Grace CPU 的性能是功耗受限的环境中的两倍。它的每瓦性能是世界上最先进 CPU 的两倍。它的数据速率是疯狂的。它是为处理超级计算机而设计的，而 Grace 是一款令人难以置信的 CPU。现在，Vera 提高了单线程性能，增加了内存容量，一切都急剧增加。这是一个巨大的芯片。这是 Vera CPU。
**English / Original**
Each one of them are completely revolutionary and the best of its kind. The Vera CPU, I'm so proud of it. In a power-constrained world, Grace CPU is two times the performance in a power-constrained world. It's twice the performance per watt of the world's most advanced CPUs. Its data rate is insane. It was designed to process supercomputers, and Grace was an incredible CPU. Now, Vera increases the single-threaded performance, increases the capacity of the memory, increases everything just dramatically. It's a giant chip. This is the Vera CPU.
## 01:04:41
**中文**
这是一个CPU。这是连接到 Rubin GPU 的。看看那个东西。这是一个巨大的芯片。现在，这是非常特别的事情，我将详细介绍这些。我想，这需要三只手，四只手才能做到这一点。好的，这就是 Vera CPU。它有88个CPU核心，并且CPU核心被设计为多线程。但是，Vera 的多线程特性经过精心设计，使得 176 个线程中的每一个都可以充分发挥其性能。所以，本质上好像有 176 个核心，但实际上只有 88 个
**English / Original**
This is one CPU. And this is connected to the Rubin GPU. Look at that thing. It's a giant chip. Now, the thing that's really special, and I'll I'll go through these. It's going to take three hands, I think, four hands to do this. Okay, so this is the Vera the Vera CPU. It's got 88 CPU cores, and the CPU cores are designed to be multithreaded. But, the multithreaded nature of of Vera was designed so that each one of the 176 threads could get its full full performance. So, it's essentially as if there's 176 cores, but only 88
## 01:05:21
**中文**
物理核心。因此，这些核心是使用一种称为空间多线程的技术来设计的。但是，IO 性能令人难以置信。这是 Rubin GPU。它的浮动性能是 Blackwell 的 5 倍。但是，重要的是要触及底线。底线是，它的晶体管数量仅为 Blackwell 的 1.6 倍。这可以告诉你一些关于当今半导体物理学水平的信息。
**English / Original**
physical cores. So, these cores were designed in in using a technology called spatial multithreading. But, the IO performance is incredible. This is the Rubin GPU. It's 5x Blackwell in floating performance. But, the important thing is go to the bottom line. The bottom line, it's only 1.6 times the number of transistors of Blackwell. That kind of tells you something about the the levels of semiconductor physics today.
## 01:05:46
**中文**
如果我们不在整个系统的基本上每个芯片的层面上进行极端的协同设计，我们怎么可能提供每年最多 1.6 倍的性能水平？因为这是你拥有的晶体管总数。即使每个晶体管的性能要高一点，比如 25%，你也不可能从所获得的晶体管数量中获得 100% 的产量。因此，1.6 倍对每年的性能提升设置了上限，除非你做了一些极端的事情，我们称之为极端协同设计。
**English / Original**
If we don't do extreme co-design at the level of basically every single chip across the entire system, how is it possible we deliver performance levels that is, you know, at best 1.6 times each year? Because that's the total number of transistors you have. And even if you were to have a little bit more um performance per transistor, say 25%, you're this impossible to get 100% yield out of the number of transistors you get. And so, 1.6x kind of puts a ceiling on how far performance can go each year unless you do something extreme, and we call it extreme co-design.
## 01:06:23
**中文**
嗯，我们所做的其中一件事就是 NVFP4 Tensor Core，这是一项伟大的发明。我们芯片内的变压器引擎不仅仅是我们放入数据路径中的四位浮点数。它是一个完整的处理器，一个处理单元，了解如何动态地、自适应地调整其精度和结构以处理不同级别的变压器，以便您可以在可能损失精度的地方实现更高的吞吐量，并在需要时恢复到尽可能高的精度。
**English / Original**
Well, one of the things that one of the things that we did, and it was a great invention, is called NVFP4 Tensor Core. The transformer engine inside our chip is not just a four-bit floating-point number somehow that we put into the data path. It is an entire processor, a processing unit that understands how to dynamically, adaptively, adjust its precision and structure to deal with different levels of the transformer so that you can achieve higher throughput wherever it's possible to lose precision and to go back to the highest possible precision wherever you need to.
## 01:07:00
**中文**
这种动态执行此操作的能力，您无法在软件中执行此操作，因为显然，它运行得太快了。因此，您必须能够在处理器内部自适应地执行此操作。这就是 NVFP4 当有人说 FP4 或 FP8 时，它对我们来说几乎毫无意义。原因是 Tensor Core 结构和所有让它发挥作用的算法。 NVFP4，我们已经发表了这方面的论文。吞吐量和精度水平能够保持的精度是完全令人难以置信的。这是一项开创性的工作。
**English / Original**
That ability to dynamically do that, you can't do this in software because, obviously, it's just running too fast. And so, you have to be able to do it adaptively inside the processor. That's what an NVFP4 When somebody says FP4 or FP8, it almost means nothing to us. And the reason for that is because it's the Tensor Core structure and all of the algorithms that makes makes it work. NVFP4, we've published papers on this already. The precision that the the level of throughput and precision is able to retain is in com- and completely incredible. This is groundbreaking work.
## 01:07:34
**中文**
如果业界希望我们将这种格式和结构成为未来的行业标准，我不会感到惊讶。这完全是革命性的。这就是我们如何能够在性能上实现如此巨大的提升，尽管我们的晶体管数量只有 1.6 倍。现在，一旦你有了一个很棒的处理节点，这是处理器节点，在里面，所以这是，例如，在这里，让我这样做。
**English / Original**
I would not be surprised if the industry would like us to make this format and this structure an industry standard in the future. This is completely revolutionary. This is how we were able to deliver such a gigantic step up in performance even though we only have 1.6 times the number of transistors. And now, once you have a great processing node, and this is the processor node, and inside, so this is this is, for example, here, let me do this.
## 01:08:07
**中文**
这是这个，哇，超级重。作为一名首席执行官，你必须保持良好的状态才能完成这项工作。好的。好的。所以，我猜这个东西可能有几百磅，我不知道。我觉得这也很有趣。快点。本来可以的。每个人都会说：“不，我不这么认为。”好的。所以，看看这个。这是最后一张。我们彻底改变了整个 MGX 底盘。
**English / Original**
This is This is wow, super heavy. You have to be a CEO in really good shape to do this job. Okay. All right. So, this thing is I'm going to guess this is probably, I don't know, couple of hundred pounds. I thought that was funny too. Come on. It could have been. Everybody's going, "No, I don't think so." All right. So, so look at this. This is the last one. We revolutionized the entire MGX chassis.
## 01:08:49
**中文**
该节点有43根电缆。零电缆。六管。 Z- 这里只有他们两个。组装这个需要两个小时。运气好的话，需要两个小时。当然，您可能会组装错误。你必须重新测试它、测试它、重新组装它。因此，组装过程非常复杂，作为我们第一台以这种方式解构的超级计算机之一，这是可以理解的。
**English / Original**
This node, 43 cables. Zero cables. Six tubes. Z- Just two of them here. It takes two hours to assemble this. If you're lucky, it takes two hours. And of course, you're probably going to assemble it wrong. You're going to have to retest it, test it, reassemble it. So, the assembly process is incredibly complicated, and it it was understandable as one of our first supercomputers that's deconstructed in this way.
## 01:09:22
**中文**
这个时间从两小时到五分钟不等。 80% 液体冷却。 100% 液体冷却。是的，真的，真的是一个突破。好的，这就是新的计算机箱，将所有这些连接到机架交换机顶部的东西（东西向流量）称为 Spectrum-X NIC。这是世界上最好的网卡。毫无疑问是NVIDIA的Mellanox，很久以前收购的Mellanox就加入了我们。嗯，他们的高性能计算网络技术是
**English / Original**
This, from two hours to five minutes. 80% liquid cooled. 100% liquid cooled. Yeah, really, really a breakthrough. Okay, so so this is the new compute chassis and what connects all of these to the top of rack switches, the east-west traffic, is called a Spectrum-X NIC. This is the world's best NIC. Unquestionably NVIDIA's Mellanox, the acquisition Mellanox that joined us a long time ago now. Um this their networking technology for high-performance computing is the
## 01:10:03
**中文**
世界上最好的，没有之一。
**English / Original**
world's best bar none.
## 01:10:05
**中文**
算法、芯片设计、所有互连、我们在其上运行的所有软件堆栈、他们的 RDMA，绝对、绝对没有，世界上最好的。现在它能够实现可编程 RDMA 和数据路径加速器，以便我们的合作伙伴（例如 AI 实验室）可以创建自己的算法，以实现他们想要如何在系统中移动数据。但这完全是世界级的 ConnectX。
**English / Original**
The algorithms, the chip design, all of the interconnects, all the software stacks that we're running on top of it, their RDMA, absolutely, absolutely bar none, the world's best. And now it has the ability to do programmable RDMA and data path accelerator so that our partners like AI labs could create their own algorithms for how they want to move data around the system. But this is completely world world class ConnectX.
## 01:10:29
**中文**
ConnectX-9 和 Vera CPU 是共同设计的，我们从未透露过它，直到 CX-9 出现才发布它，因为我们为新型处理器共同设计了它。您知道，ConnectX-9 或 CX-8 和 Spectrum-X 彻底改变了以太网在人工智能领域的应用方式。人工智能的以太网流量更加密集，需要更低的延迟。流量的瞬时激增与以太网所见不同，因此我们创建了 Spectrum-X，即 AI 以太网。
**English / Original**
ConnectX-9 and the Vera CPU were co-designed and we never revealed it not never never released it until CX-9 came along because we we co-designed it for a new type of processor. You know, ConnectX-9 or CX-8 and Spectrum-X revolutionized how Ethernet was done for artificial intelligence. Ethernet traffic for AI is much, much more intense, requires much lower latency. The the instantaneous surge of traffic is unlike anything Ethernet sees and so we created Spectrum-X which is AI Ethernet.
## 01:11:07
**中文**
2 年前，我们发布了 Spectrum-X。如今，NVIDIA 是世界上最大的网络公司。所以它非常成功并被用于许多不同的安装中。它正在席卷人工智能领域。性能令人难以置信，特别是当您拥有 200 兆瓦数据中心或千兆瓦数据中心时，这些都是数十亿美元。假设一个千兆瓦数据中心价值 500 亿美元。如果网络性能允许您额外提供 10%，那么在 Spectrum-X 的情况下提供 25% 更高的吞吐量并不罕见。如果它
**English / Original**
2 years ago we announced Spectrum-X. NVIDIA today is the largest networking company the world has ever seen. So it's been so successful and used in so many different installations. It is just sweeping the AI landscape. The performance is incredible, especially when you have a 200 megawatt data center or if you have a gigawatt data center, these are billions of dollars. Let's say a gigawatt data center is 50 billion dollars. If the networking performance allows you to deliver an extra 10% in the case of Spectrum-X delivering 25% higher throughput is not uncommon. If it
## 01:11:48
**中文**
只交付 10%，价值 50 亿美元。网络是完全免费的。这就是为什么每个人都使用 Spectrum-X 的原因。这真是一件令人难以置信的事情。现在我们要发明一种新的数据处理类型。因此 Spectrum-X 适用于东西向流量。我们现在有了一款名为 BlueField-4 的新处理器。 BlueField-4 允许我们将一个非常大的数据中心隔离其中的不同部分，以便不同的用户可以使用其中的不同部分，确保如果他们决定虚拟化，一切都可以虚拟化。因此，您可以卸载大量虚拟化软件、安全软件、网络软件
**English / Original**
were to just deliver 10%, that's worth 5 billion dollars. The networking is completely free. Which is the reason why well everybody uses Spectrum-X. It's just an incredible thing. And now we're going to invent a new type a new type of data processing. And so Spectrum-X is for east-west traffic. We now have a new processor called BlueField-4. BlueField-4 allows us to take a large large very large data center isolate different parts of it so that different users could use different parts of it, make sure that everything could be virtualized if they decide to be virtualized. So you offload a lot of the virtualization software, the security software, the networking software for
## 01:12:30
**中文**
你的南北交通。因此，BlueField-4 是每一个计算节点的标准配置。 BlueField-4 还有第二个应用程序，我稍后将讨论它。这是一款革命性的处理器，我对此感到非常兴奋。这是 NVLink 6 开关，就在这里。这是这个开关，这里的开关，NVLink 开关里面有四个。
**English / Original**
your north-south traffic. And so BlueField-4 comes standard with every single one of these compute nodes. BlueField-4 has a second application I'm going to talk about in just a second. This is a revolutionary processor and I'm so excited about it. This is the NVLink 6 switch and it's right here. This is the this switch this switch here, there's four of them inside the NVLink switch here.
## 01:13:00
**中文**
这些交换芯片中的每一款都具有历史上最快的 SerDes。世界还勉强达到 200 吉比特。这是每秒 400 吉比特的交换机。这之所以如此重要，是因为我们可以让每个 GPU 在同一时间与所有其他 GPU 进行通信。这个交换机位于这些机架之一的背板上，使我们能够传输相当于全球互联网数据量两倍的数据。
**English / Original**
Each one of these switch chips has the fastest 30s in history. The world is barely getting to 200 gigabits. This is 400 gigabits per second switch. The reason why this is so important is so that we could have every single GPU talk to every other GPU at exactly the same time. This switch this switch on the backplane of one of these racks enables us to move the equivalent of twice the amount of the global internet data.
## 01:13:33
**中文**
全球所有互联网数据的速度都是原来的两倍。整个地球互联网的横截面带宽约为每秒 100 TB。即每秒 240 TB。这样就可以正确地看待它。这样每个 GPU 都可以在同一时间与其他每个 GPU 一起工作。
**English / Original**
All of the world's internet data at twice the speed. You take the cross-sectional bandwidth of the entire planet's internet, it's about 100 terabytes per second. This is 240 terabytes per second. So that puts it in perspective. This is so that every single GPU can work with every single other GPU at exactly the same time.
## 01:13:53
**中文**
好的？然后在那之上。好的，这是一个机架。这是一个架子。每个机架，正如你所看到的，这一机架中的晶体管数量是 1.7 倍 是的，你能帮我做一下吗？所以这通常是 2 吨左右，但今天是 2 又 1/2 吨，因为嗯，当他们运送它时，他们忘记将水排出。
**English / Original**
Okay? Then on top of that on top of that. Okay, so this is one rack. This is one rack. Each one of the racks, as you could see, the number of transistors in this one rack is 1.7 times Yeah, could you do this for me? So this is it's usually about 2 tons but today is 2 and 1/2 tons because um when they shipped it they forgot to drain the water out of it.
## 01:14:27
**中文**
所以我们从加利福尼亚运了很多水。你能听到它尖叫吗？您知道，当您旋转 2 又 1/2 吨时，您会发出一点尖叫声。这背后是 NVLink 的脊椎。基本上是 2 英里的铜缆。铜是我们所知的最好的导体，这些都是屏蔽铜电缆、结构化铜电缆，是世界上计算系统中使用最多的电缆。
**English / Original**
So we we shipped a lot of water from California. Can you hear it squealing? You know, when you're rotating 2 and 1/2 tons you're going to squeal a little. So behind behind this are the NVLink spines. Basically 2 miles of copper cables. Copper is the best conductor we know and these are all shielded copper cables, structured copper cables, the most the world's ever used in computing systems ever.
## 01:14:58
**中文**
我们的 SerDes以每秒 400 吉比特的速度将铜缆从机架顶部一直驱动到机架底部。太不可思议了。因此，铜缆总长 2 英里，共 5,000 根铜缆。这使得 NVLink 主干成为可能。这是真正开启 MGX 系统的革命。
**English / Original**
And and our 30s drive the copper cables from the top of the rack all the way to the bottom of the rack at 400 gigabits per second. It's incredible. And so this has 2 miles of total copper cables, 5,000 copper cables. And this makes the NVLink spine possible. This is the revolution that that really started the MGX system.
## 01:15:23
**中文**
现在我们决定创建一个行业标准体系，以便整个生态系统、我们所有的供应链都可以对这些组件进行标准化。这些 MGX 系统由大约 8 万个不同的组件组成，如果我们每年都更换它，那就完全是浪费。从富士康到广达再到纬创资通，每一家主要的计算机公司，你知道，这个名单不胜枚举，再到惠普、戴尔和联想，每个人都知道如何构建这些系统。事实上，我们可以将 Vera Rubin 挤进这个，即使性能如此之多
**English / Original**
Now we we decided that we would create an industry standard system so that the entire ecosystem, all of our supply chain could standardize on these components. There's some 80 thousand different components that make up this these MGX systems and it's a total waste if we were to change it every single year. Every single major computer company from Foxconn to Quanta to Wistron, you know, the list goes on and on and on to HP and Dell and Lenovo, everybody knows how to build these systems. And so the fact that we could squeeze Reuben Vera Reuben into this even though the performance is so much
## 01:16:04
**中文**
如此之高，非常重要的是，功率是原来的两倍。维拉·鲁本的力量是Grace Hopper的两倍。然而，进入其中的空气、气流几乎是一样的，这就是奇迹。非常重要的是，进入其中的水温度相同，均为 45°C。数据中心的温度为 45°C，无需冷水机。
**English / Original**
so much higher and very importantly the power is twice as high. The power Vera Reuben is twice as high as Grace Hopper. And yet and this is the miracle the air that goes into it, the the air flow is about the same. And very importantly the water that goes into it is the same temperature, 45°C. With 45°C no water chillers are necessary for data centers.
## 01:16:33
**中文**
我们基本上是用热水来冷却这台超级计算机。它是如此的高效。这就是新的新机架。晶体管数量增加了 1.7 倍，但峰值推理性能增加了五倍，峰值训练性能增加了三倍半。它们通过 Spectrum-X 在顶部连接。噢，谢谢。
**English / Original**
We're basically cooling this supercomputer with hot water. It's so incredibly efficient. And so this is this is the new the new rack. 1.7 times more transistors but five times more peak inference performance, three and a half times more peak um uh uh training performance. They're connected on top using Spectrum-X. Oh, thank you.
## 01:17:06
**中文**
这是世界上第一个使用台积电新工艺制造的芯片，我们共同创新的称为 Co-op。它是一种硅光子集成硅光子工艺技术。这使我们能够将硅光子学直接应用到芯片上。这是 512 个端口，速率为每秒 200 吉比特。这就是新型以太网 AI 交换机 Spectrum-X 以太网交换机。看看这个巨大的芯片。但真正令人惊奇的是它直接连接了硅光子学。
**English / Original**
This is this is the world's first manufacturing chip using uh TSMC's new process that we co-innovated called Co-op. It's a silicon photonics integrated silicon photonics process technology. And this allows us to take silicon photonics directly right to the chip. And this is 512 ports at 200 gigabits per second. And this is the new Ethernet AI switch, the Spectrum-X Ethernet switch. And look at this giant chip. But what's really amazing it's got silicon photonics directly connected to it.
## 01:17:42
**中文**
激光进来了。激光从这里进来。激光从这里射进来。光学器件就在这里，它们连接到数据中心的其余部分。我稍后会向您展示，但它位于机架顶部，这是新的 Spectrum-X 硅光子开关。好的？我想告诉你一些新的事情。正如我几年前提到的那样，我们推出了 Spectrum-X，以便我们可以重新发明网络的完成方式。
**English / Original**
And lasers come in. Lasers come in through here. Lasers come in through here. The optics are here and they connect out to the rest of the data data center. This I'll show you in a second but this is on top of the rack and this is the new Spectrum-X silicon photonics switch. Okay? And we have something new I want to tell you about. So just as I mentioned a couple years ago we introduced Spectrum-X so that we could reinvent the way that networking is done.
## 01:18:19
**中文**
嗯，以太网确实很容易管理，每个人都有一个以太网堆栈，世界上每个数据中心都知道如何处理以太网。嗯，我们当时唯一使用的是 InfiniBand，它用于超级计算机。 InfiniBand 的延迟非常低，但当然，InfiniBand 的软件堆栈和整个可管理性对于使用以太网的人来说非常陌生。因此我们决定首次进入以太网交换机市场。正如我提到的，Spectrum-X 刚刚起飞，它使我们成为世界上最大的网络公司。下一代 Spectrum-X 将继承这一传统。
**English / Original**
Um Ethernet is really easy to manage and everybody has an Ethernet stack and every data center in the world knows how to deal with Ethernet. Um and the only thing that we were we were using at the time was called InfiniBand which is used for supercomputers. InfiniBand is very low latency um but of course the software stack, the entire manageability of InfiniBand is very alien to the people who use Ethernet. So we decided to enter the Ethernet switch market for the very first time. Spectrum-X that just took off and it made us the largest networking company in the world as I mentioned. This this next generation Spectrum-X is going to carry on that tradition.
## 01:18:57
**中文**
但正如我之前所说，人工智能已经重塑了整个计算堆栈。计算堆栈的每一层，理所当然地认为，当人工智能开始在全球企业中部署时，它也将重塑存储的完成方式。嗯，人工智能不使用 SQL。人工智能使用语义信息。当人工智能被使用时，它会创建这种临时知识，临时内存，称为KV 缓存。
**English / Original**
But just as I said earlier, AI has reinvented the whole computing stack. Every layer of the computing stack, it stands to reason that when AI starts to get deployed in the world's enterprises, it's going to also reinvent the way storage is done. Well, AI doesn't use SQL. AI use semantics information. And when AI is being used, it creates this temporary knowledge, temporary temporary memory called KB cache.
## 01:19:24
**中文**
键值组合，不过是KV 缓存。基本上，人工智能的缓存，人工智能的工作内存。 AI的工作记忆存储在HBM记忆中。 GPU 在模型（整个模型）中读取的每个 token的每个 token。它读取整个工作内存并生成一个token。并将该token存储回 KV 缓存中。然后下一次它这样做时，它会读取整个内存，读取它并通过我们的 GPU 流式传输，然后生成另一个token。好吧，它会重复地执行此操作，一个又一个token，显然，如果您有一个很长的
**English / Original**
Key value combination, but it's a KB cache. Basically, the cache of the AI, the working memory of the AI. And the working memory of the AI is stored in the HBM memory. Every single token for every single token the GPU reads in the model, the entire model. It reads in the entire working memory and it produces one token. And it stores that one token back into the KB cache. And then the next the next time it does that, it reads in the entire memory, reads it and it streams it through our GPU and then generates another token. Well, it does this repeatedly token after token after token and obviously, if you have a long
## 01:20:07
**中文**
与人工智能对话，随着时间的推移，记忆、上下文记忆将会极大地增长。更不用说模型在不断增长，我们使用人工智能的次数也在增加。我们希望这个人工智能陪伴我们一生，并记住我们与它进行的每一次对话。
**English / Original**
conversation with that AI, over time that memory, that context memory is going to grow tremendously. Not to mention the models are growing, the number of turns that we're using the AIs are are are increasing. We would like to have this AI stay with us our entire life and remember every single conversation we've ever had with it.
## 01:20:24
**中文**
正确的？我要求的每一项研究。当然，随着共享超级计算机的人数将继续增长。因此，最初适合 HBM 的上下文内存不再足够大。去年，我们创造了 Grace Blackwell 的非常快的记忆，我们称之为快速上下文记忆。这就是我们将Grace直接与Hopper联系起来的原因。这就是为什么我们将 Grace 直接连接到 Blackwell，以便我们可以扩展上下文记忆。但即便如此还不够。所以，下一个解决方案当然是进入网络，南北网络
**English / Original**
Right? Every single lick of research that I've asked it for. Of course, with the number of people that will be sharing the supercomputer is going to continue to grow. And so, this context memory which started out fitting inside an HBM is no longer large enough. Last year, we created Grace Blackwell's very fast memory we called fast context memory. In That's the reason why we connected Grace directly to Hopper. That's why we connected Grace directly to Blackwell so that we can expand the context memory. But even that is not enough. And so, the next solution, of course, is to go off onto the network, the north-south network
## 01:21:04
**中文**
送到公司仓库。
**English / Original**
off to the storage of the company.
## 01:21:06
**中文**
但如果同时运行大量人工智能，该网络将不再足够快。所以，答案非常明确，那就是采取不同的做法。因此，我们创建了 BlueField 4，这样我们基本上就可以在机架中拥有非常快速的 KV 缓存上下文内存存储。因此，我将在短短 1 秒内向您展示，但是存在一种全新的存储系统类别，并且整个行业非常兴奋，因为这对于当今进行大量token生成的几乎每个人、人工智能实验室、云服务提供商来说都是一个痛点。他们确实受到网络数量的影响
**English / Original**
But if you have a whole lot of AIs running at the same time, that network is no longer going to be fast enough. So, the answer is very clearly to do it different. And so, we we created Bluefield 4 so that we could essentially have a very fast KB cache context memory store right in the rack. And so, I'll show you in just 1 second, but there's a whole new category of storage systems and the industry is so excited because this is a pain point for just about everybody who does a lot of token generation today, the AI labs, the cloud service providers. They're really suffering from the amount of network
## 01:21:47
**中文**
造成的流量是由 KV 缓存移动引起的。因此，我们创建一个新平台、一个新处理器来运行整个 Dynamo KV 缓存上下文内存管理系统并将其非常靠近机架的其余部分的想法是完全革命性的。所以，就是这样。这是它就坐在这里。
**English / Original**
traffic that's causing being caused by KB cache moving around. And so, the idea that we would create a new platform, a new processor to run the entire Dynamo KB cache context memory management system and to put it very close to the rest of the rack is completely revolutionary. So, this is it. This is it sits right here.
## 01:22:11
**中文**
所以，这就是所有计算节点。其中每一个都是 NVLink 72。因此，这是 Vera Rubin NVLink 72。144 Rubin GPU。这是存储在这里的上下文内存。每一个背后都有四个BlueField。每个 BlueField 背后都有 150 TB 的内存上下文内存。对于每个 GPU，一旦分配完毕，每个 GPU 将获得额外的 16 TB。
**English / Original**
So, this this is all the compute nodes. Each one of these is NVLink 72. So, this is Vera Rubin NVLink 72. 144 Rubin GPUs. This is the context memory that's stored here. Behind each one of these are four Bluefields. Behind each Bluefield is 150 terabytes of memory context memory. And for each GPU, once you allocate it across, each GPU will get an additional 16 terabytes.
## 01:22:43
**中文**
现在，在这个节点内，每个 GPU 基本上有 1 TB。现在，通过此后备存储直接在相同的东西向流量上以完全相同的数据速率（每秒 200 GB）跨越该计算节点的整个结构，您将获得额外的 16 TB 内存。好的？这是管理平面。
**English / Original**
Now, inside this node, each GPU essentially has 1 terabyte. And now, with this backing store here directly on the same east-west traffic at exactly the same data rate, 200 gigabits per second across literally the entire fabric of this compute node, you're going to get an additional 16 terabytes of memory. Okay? And this is the management plane.
## 01:23:11
**中文**
这些是将它们连接在一起的 Spectrum-X 交换机。在这里，末端的这些交换机将它们连接到数据中心的其余部分。好的？这就是Vera Rubin。现在，有几件事确实令人难以置信。因此，我提到的第一件事是，整个系统的能源效率提高了一倍，本质上是温度性能的两倍，即即使功率高一倍，所使用的能源量也高一倍，
**English / Original**
These are these are the Spectrum-X switches that connects all of them together. And over here these switches at the end connects them to the rest of the data center. Okay? And so, this is the Vera Rubin. Now, there's several things that's really incredible about it. So, the first thing that I mentioned is that the this entire system is twice the energy efficiency, essentially the the the temperature performance in the sense that that even though the power is twice as high, the amount of energy used is twice as high, the amount of
## 01:23:54
**中文**
计算结果比这个高很多倍，但进入其中的液体仍然是45°C。这使我们能够节省全球数据中心约 6% 的电力。所以，这是一件大事。第二件大事是整个系统现在是机密计算安全的。这意味着，一切都在传输、静止和计算过程中进行编码。现在每条总线都已加密。
**English / Original**
computation is many times higher than that, but the liquid that goes into it still 45°C. That enables us to save about 6% of the world's data center power. So, that's a very big deal. The second very big deal is that this entire system is now confidential computing safe. Meaning, everything is encoded in transit, at rest and during compute. And every single bus is now encrypted.
## 01:24:23
**中文**
每一个 PCI Express、每一个 NVLink、每一个你知道的、CPU 和 GPU 之间、GPU 到 GPU 之间的 NVLink，现在一切都已加密。因此，它是机密计算的安全。这让公司感到安全，因为他们的模型正在被其他人部署，但其他人永远不会看到。好的？因此，这个特殊的系统不仅具有令人难以置信的能源效率，而且还有另一件事令人难以置信。
**English / Original**
Every PCI Express, every NVLink, every you know, NVLink between CPU and GPU, between GPU to GPU, everything is now encrypted. And so, it's confidential computing safe. This allows companies to feel safe that their models are being deployed by somebody else, but it will never be seen by anybody else. Okay? And so, this particular system is not only incredibly energy efficient, but there's one other thing that's incredible.
## 01:24:55
**中文**
由于人工智能工作负载的性质，它会在称为 allreduce 的计算层中瞬间激增。同时使用的电流量和能量量确实是超乎想象的。很多时候，它们会飙升 25%。我们现在对整个系统进行功率平滑，这样您就不必超额配置 25% 的供电能力，或者如果您超额配置 25% 的供电能力，您就不必浪费或未使用 25% 的能源。现在，您可以填满整个功率预算，并且无需再进行额外配置。
**English / Original**
Because of the nature of the workload of AI, it spikes instantaneously with this computation layer called all reduce. The amount of current, the amount of energy that is used simultaneously is really off the charts. Often times, they'll spike up 25%. We now have power smoothing across the entire system so that you don't have to over provision by 25 times or if you over provision by 25 times, you don't have to leave 25% of the energy um squandered or unused. And so now, you could fill up the entire power budget and you don't have to provision beyond that.
## 01:25:35
**中文**
当然，最后一件事是性能。那么，我们来看看它的表现吧。这些只是构建人工智能超级计算机的人会喜欢的图表。我们需要每一个芯片，对每一个系统进行彻底的重新设计，并重写整个堆栈，才使这一切成为可能。
**English / Original**
And then the last thing, of course, is performance. So, let's take a look at the performance of this. These are only charts that people who build AI supercomputers would love. It took every single one of these chips, complete redesign of every single one of the systems and rewriting the entire stack for us to make this possible.
## 01:25:52
**中文**
基本上，这是训练人工智能模型，这是第一列。训练人工智能模型的速度越快，您就能更快地将下一个前沿领域推向世界。这是您的上市时间。这就是技术领先。这就是你的定价能力。因此，就绿色而言，这本质上是一个 10 万亿参数模型。我们从 DeepSeek 对其进行了扩展。
**English / Original**
Basically, this is training the AI model, this first column. The faster you train AI models the faster you can get the next frontier out to the world. This is your time to market. This is technology leadership. This is your pricing power. And so, in the case of the green, this is essentially a 10 trillion parameter model. We scaled it up from Deep Seek.
## 01:26:16
**中文**
DeepSeek 这就是为什么我们称其为 DeepSeek++。在 100 万亿个token上训练 10 万亿参数模型。好的？这就是我们对构建下一个前沿模型所需的模拟预测。下一个前沿模型呃 Elon 已经提到了 Grok 的下一个版本，Grok 5，我认为是 7 万亿个参数。所以，这是 10。
**English / Original**
Deep Seek That's why we call it Deep Seek plus plus. A training a 10 trillion parameter model on 100 trillion tokens. Okay? And that's this is our simulation projection of what it would take for us to build the next frontier model. The next frontier model uh Elon's already mentioned that the next version of Grok, Grok 5, I think is 7 trillion parameters. So, this is 10.
## 01:26:38
**中文**
和 和 绿色的是布莱克威尔。在这里，以 um uh Rubin 为例，请注意吞吐量要高得多，因此，在我们在这里给出的时间（即 1 个月）内训练模型只需要这些系统的 1/4。好的？所以时间对每个人来说都是一样的。现在，你能以多快的速度训练该模型以及你能训练多大的模型是你首先到达前沿的方式。第二部分是工厂的吞吐量。
**English / Original**
And and and the green is Blackwell. And here, in the case of um uh Rubin, notice the throughput is so much higher and therefore, it only takes 1/4 as many of these systems in order to train the model in the time that we gave it here, which is 1 month. Okay? And so time time is the same for everybody. Now, how much how fast you can train that model and how large a model you can train is how you're going to get to the frontier first. The second part is your factory throughput.
## 01:27:13
**中文**
布莱克威尔又变绿了，工厂的吞吐量很重要，因为你的工厂在一吉瓦的情况下是 500 亿美元。价值 500 亿美元的数据中心只能消耗 1 吉瓦的电力。因此，如果您的每瓦吞吐量性能非常好而不是非常差，那么这将直接转化为您的收入。您的数据中心的收入与第二列直接相关。就布莱克威尔而言，这个数字大约是Hopper的 10 倍。就Rubin而言，大约是 10
**English / Original**
Blackwell is green again and factory throughput is important because your factory is in the case of a gigawatt is $50 billion. A $50 billion data center can only consume 1 gigawatt of power. And so, if your performance your throughput per watt is very good versus quite poor, that directly translates to your revenues. Your revenues of your data center is directly related to the second second column. And in the case of Blackwell, it was about 10 times over Hopper. In the case of Rubin, it's going to be about 10
## 01:27:53
**中文**
再次高出几倍。
**English / Original**
times higher again.
## 01:27:55
**中文**
好的？就现在的情况而言，token的成本，生成token的成本效益如何，大约是 Rubin 的 1/10。就像是的情况一样。所以，这就是我们如何让每个人进入下一个前沿，将人工智能推向新的水平。当然，要高效、经济地建设这些数据中心。
**English / Original**
Okay? And in the case of now, the um the cost of the tokens, how cost-effectively it is to generate the token this is Rubin about 1/10. Just as in the case of Yep. So, that's how This is how we're going to get everybody to the next frontier to um push AI to the next level. And of course, to build these data centers energy efficiently and cost efficiently.
## 01:28:30
**中文**
所以，就是这样。 This is NVIDIA today.你知道，我们提到我们制造芯片，但正如你所知，NVIDIA 现在制造整个系统。 AI是一个全栈。我们正在重塑人工智能的各个领域，从芯片到基础设施，从模型到应用程序。我们的工作是创建整个堆栈，以便所有人都可以为世界其他地方创建令人难以置信的应用程序。谢谢大家的到来。祝您度过一个精彩的 CES。
**English / Original**
So, this is it. This is NVIDIA today. You know, we mentioned that we build chips, but as you know, NVIDIA builds entire systems now. And AI is a full stack. We we're reinventing AI across everything from chips to infrastructure to models to applications. And our job is to create the entire stack so that all of you could create incredible applications uh for the rest of the world. Thank you all for coming. Have a great CES.
## 01:29:01
**中文**
现在在我让你们走之前之前，呃，我们必须把一大堆幻灯片留在剪辑板上。因此，我们在这里有一些删减。我想这对你来说会很有趣。伙计们，祝你有个愉快的 CES。 >> 然后切割。 NVIDIA 在 CES 现场拿下四项。标记。吊杆麦克风。行动。对不起各位。平台迁移吧？
**English / Original**
Now Before Before I Before I let you guys go, uh there were a whole bunch of slides we have to we have to leave on the cutting floor. And so, we have some outtakes here. I think it'll be fun for you. Have a great CES, guys. >> And cut. Nvidia live at CES take four. Marker. Boom mic. Action. Sorry guys. Platform shift, huh?
## 01:29:41
**中文**
应该可以做到这一点。让我们滚动相机。绿色的阴影。明亮快乐的绿色。世界上最强大的人工智能超级计算机，您可以将其插入我的烤面包机旁边的墙上。嘿伙计们，我又被困住了。我很抱歉。 >> 这张幻灯片永远行不通。让我们把它剪掉吧。你好？你能听到我吗？就像我说的，路由器，因为并不是每个问题都需要最大、最智能的模型。
**English / Original**
That should do it. And let's roll camera. A shade of green. A bright happy green. World's most powerful AI supercomputer you can plug into the wall next to my toaster. Hey guys, I'm I'm stuck again. I'm so sorry. >> This slide is never going to work. Let's just cut it. Hello? Can you hear me? So like I was saying, the router because not every problem needs the biggest, smartest model.
## 01:30:17
**中文**
恰到好处。不，不，不要失去任何一个。这个新的六芯片 Rubin 平台打造出了一台令人惊叹的人工智能超级计算机。就这样吧，小家伙。哦，不。不，不是缩放定律。车上有一只松鼠。准备好让松鼠离开。轻轻地请松鼠走开。您知道当今最好的模特都是专家的混合体吗？
**English / Original**
Just the right one. No. No. Don't lose any of them. This new six-chip Rubin platform makes one amazing AI supercomputer. There you go, little guy. Oh, no. No, not the scaling laws. There is a squirrel on the car. Be ready to make the squirrel go away. Ask the squirrel gently to move away. Did you know the best models today are all mixture of experts?
## 01:30:48
**中文**
嘿。大家都去哪儿了？
**English / Original**
Hey. Where'd everybody go?