-
The chip offers significant performance improvements of 60% to 90% compared to the current H100 model for AI inference.
-
H200-powered systems will be available from Nvidia’s hardware partners and major cloud service providers.
-
Nvidia’s GPUs have become highly sought after for training AI models, resulting in product shortages.
-
Nvidia’s new strategy of more frequent, higher performance product launches poses a challenge for rivals like AMD.