The newly announced NVIDIA’s DGX GH200 AI supercomputer delivers 1 exaflop of performance for generative AI.
Since the release of OpenAI's ChatGPT, Nvidia chips have high exposure to generative AI.
This new class of large-memory AI supercomputer is powered by NVIDIA® GH200 Grace Hopper Superchips and the NVIDIA NVLink® Switch System — created to enable the development of giant, next-generation models for generative AI language applications, recommender systems and data analytics workloads.
The massive shared memory space of DGX GH200 uses NVLink interconnect technology with the NVLink Switch System to combine 256 GH200 superchips, allowing them to perform as a single GPU. This provides 1 exaflop of performance and 144 terabytes of shared memory — nearly 500x more memory than the previous generation NVIDIA DGX A100, which was introduced in 2020.
That lets developers build large language models for generative AI chatbots, complex algorithms for recommender systems, and graph neural networks used for fraud detection and data analytics.
Since the release of OpenAI's ChatGPT, Nvidia chips have high exposure to generative AI.
This new class of large-memory AI supercomputer is powered by NVIDIA® GH200 Grace Hopper Superchips and the NVIDIA NVLink® Switch System — created to enable the development of giant, next-generation models for generative AI language applications, recommender systems and data analytics workloads.
The massive shared memory space of DGX GH200 uses NVLink interconnect technology with the NVLink Switch System to combine 256 GH200 superchips, allowing them to perform as a single GPU. This provides 1 exaflop of performance and 144 terabytes of shared memory — nearly 500x more memory than the previous generation NVIDIA DGX A100, which was introduced in 2020.
That lets developers build large language models for generative AI chatbots, complex algorithms for recommender systems, and graph neural networks used for fraud detection and data analytics.
 |
Google Cloud, Meta and Microsoft are among the first expected to gain access to the DGX GH200, which can be used as a blueprint for future hyperscale generative AI infrastructure.
The GH200 AI Superchip, which Nvidia CEO, HUANG, said is now in full production, combines an energy-efficient NVIDIA Grace CPU with a high-performance NVIDIA H100 Tensor Core GPU in one superchip.
Interestingly, NVIDIA is also building its own massive AI supercomputer called – NVIDIA Helios – which is expected to come online this year. NVIDIA Helios, the supercomputer will feature four DGX GH200 systems. Each will be interconnected with NVIDIA Quantum-2 InfiniBandnetworking to supercharge data throughput for training large AI models.
Nvidia CEO announced more than 400 system configurations that are coming to market powered by NVIDIA’s latest Hopper, Grace, Ada Lovelace and BlueField architectures. The goal is to tackle the most complex challenges in AI, data science and high performance computing.
The GH200 AI Superchip, which Nvidia CEO, HUANG, said is now in full production, combines an energy-efficient NVIDIA Grace CPU with a high-performance NVIDIA H100 Tensor Core GPU in one superchip.
Interestingly, NVIDIA is also building its own massive AI supercomputer called – NVIDIA Helios – which is expected to come online this year. NVIDIA Helios, the supercomputer will feature four DGX GH200 systems. Each will be interconnected with NVIDIA Quantum-2 InfiniBandnetworking to supercharge data throughput for training large AI models.
Nvidia CEO announced more than 400 system configurations that are coming to market powered by NVIDIA’s latest Hopper, Grace, Ada Lovelace and BlueField architectures. The goal is to tackle the most complex challenges in AI, data science and high performance computing.
Advertisements