9621 Agnes Crossing, Lake Suzanneview, New Mexico Island 84604-9295.
The American company Nvidia presented a new graphics processor H200, designed for training generative artificial intelligence models. CNBC reports this, according to Ukrinform. The new GPU is an
CNBC reports this, according to Ukrinform.
The new GPU is an upgrade from the previous H100 chip, which OpenAI used to train its state-of-the-art language model GPT-4. H100 chips cost between $25,000 and $40,000.
The most important development of the H200 is that it has 141 GB new generation HBM3 memory.
Nvidia also said the H200 will produce data throughput almost twice as fast as the H100. This conclusion was reached through test inputs at the company using Meta company’s Llama 2 LLM language model.
H200, which is expected to go on sale in the second quarter of 2024, will compete with AMD’s MI300X GPU.
The AMD chip is said to be similar to the H200 but with additional memory compared to its predecessors.
Nvidia added that the H200 will be compatible with the H100, meaning companies won’t need to change their server systems or software to use the new version of the chip.
As reported by Ukrinform, the British company Vodafone announced its intention to create a special chipset architecture for Open RAN technology together with the American company Intel.
Photo: Walid Berrazeg | Light Rocket | Getty Images
Source: Ukrinform
As an experienced journalist and author, Mary has been reporting on the latest news and trends for over 5 years. With a passion for uncovering the stories behind the headlines, Mary has earned a reputation as a trusted voice in the world of journalism. Her writing style is insightful, engaging and thought-provoking, as she takes a deep dive into the most pressing issues of our time.
