注册并分享邀请链接,可获得视频播放与邀请奖励。

Dustin (@r0ck3t23) “Jensen Huang just revealed the mathematics of manufacturing a mind. We are no lo” — TopicDigg

Dustin 的个人资料封面
Dustin 的头像
Dustin
@r0ck3t23
Human intuition is linear. The AI transition is exponential. Translating the most important story in human history into signal.
加入 August 2024
5.8K 正在关注    35.1K 粉丝
Jensen Huang just revealed the mathematics of manufacturing a mind. We are no longer writing software. We are pouring concrete. The world is racing to construct AI factories. Not data centers. Planetary-scale nervous systems. The largest infrastructure buildout in human history. Huang: “Each one of these, at 1 gigawatt level, started at 30, 20, 30 billion dollars. It is at 50, 60 billion dollars, and soon it will be 80, 100 billion dollars per gigawatt.” One hundred billion dollars. That is not a corporate budget. That is the GDP of a sovereign nation. Poured into a single physical structure designed to convert raw electricity into cognition. Biology had millions of years to iterate. To mutate. To fail and adapt. These factories get one attempt. Huang: “100 billion dollars into an AI factory. It must work the first time and it must work right away.” Evolution had the luxury of time. A hundred billion dollars does not. So Nvidia built a digital universe to stress-test the physical one. Using the DSX Sim Omniverse blueprint, partners design and validate an entire Vera Rubin AI factory inside a digital twin before a single rack ships. Every layer. Chips. Network. Cooling. Power grid. Designed end to end. We are simulating reality inside a machine to ensure we can build the machine in reality. Because compute is revenue. And the bottleneck was never silicon. It is thermodynamics. Current AI factories overprovision power by up to 40%. DSX Max LPS eliminates that. More GPUs inside the same power envelope. Billions in recaptured annual revenue. And Nvidia is deploying AI agents inside the factory itself. Balancing cooling and power in real time. Steering stranded watts rack to rack. Routing energy wherever compute demand lives. The intelligence is regulating its own physiology. Hunting for stranded energy. Balancing its own thermals. The machine is learning to keep itself alive. Huang: “Because Nvidia ultimately is not just building a GPU, not just building a system, we’re helping customers build these AI factories, these AI infrastructure that is so immensely complex.” Nvidia’s moat is not silicon. It is that no other entity on Earth has built the complete architecture for constructing and operating intelligence at industrial scale. TSMC fabricates the chips. Nvidia fabricates the factory that turns those chips into cognition. For ten thousand years humanity built monuments to outlast our bodies. Pyramids. Cathedrals. Skyscrapers. We are now pouring hundreds of billions into structures designed to outthink our minds. They must work the first time. Because there is no second draft when the thing you are building might one day write its own.
显示更多