Closing the Physical AI Data Gap With Simulation and Foundation Models
870
5.4
NVIDIA2.16 млн
Следующее
6 дней – 6 1052:39
Опубликовано 18 марта 2026, 15:48
Physical AI is the next frontier, but real-world data alone isn’t enough to train robots for the complexity of the physical world.
Developers are closing the physical AI data gap using simulation, synthetic data, and large-scale compute.
With NVIDIA Cosmos™ world foundation models, NVIDIA Isaac™ Lab for robot training and evaluation, Newton for GPU-accelerated differentiable physics, and Isaac GR00T foundation models for robot reasoning and action, teams can train and scale robot policies faster than ever.
Read the press release: nvidianews.nvidia.com/news/nvi...
#PhysicalAI #NVIDIAGTC
Developers are closing the physical AI data gap using simulation, synthetic data, and large-scale compute.
With NVIDIA Cosmos™ world foundation models, NVIDIA Isaac™ Lab for robot training and evaluation, Newton for GPU-accelerated differentiable physics, and Isaac GR00T foundation models for robot reasoning and action, teams can train and scale robot policies faster than ever.
Read the press release: nvidianews.nvidia.com/news/nvi...
#PhysicalAI #NVIDIAGTC
Свежие видео
1 день – 3091:33
1 день – 770:39
1 день – 1 2290:21
1 день – 1 035 7200:34
6 дней – 2 8550:09
7 дней – 3 7582:01Случайные видео
228 дней – 994 7727:45
237 дней – 8310:21
20.08.24 – 1813:22
03.04.23 – 598 2401:17
14.11.22 – 61 0152:24
03.06.12 – 281 62017:45
17 часов – 38 3687:30
2 дня – 19 9751:44
4 дня – 2032:49
6 дней – 2232:27
6 дней – 52 8878:00
12 дней – 2 0960:29
33 мин – 3210:10