Nvidia announced a new AI chip with triple the memory for next year (2024).
The GH200 will be able to run the most complex generative AI models, such as large language models, recommender systems and vector databases.
Nvidia, the dominant player in providing high-end processors for generative AI applications, will launch a new chip that will be even more powerful as the demand for large AI models grows.
The company unveiled the GH200 super chip, which Nvidia said can handle “the most complex generative AI workloads, spanning large language models, recommender systems and vector databases.”
The GH200 will have the same GPU as the H100, Nvidia’s current flagship and most popular AI product, but triple the memory capacity. The company said systems running on GH200 will start in the second quarter of 2024.
Nvidia did not disclose the price for the GH200; the H100 line currently sells for around $40,000.
Complex AI models need powerful GPUs to perform the calculations necessary to generate text or a photo of a horse in the style of Banksy. Running these models requires a lot of processing power, and some have to “split” the models among other GPUs just to run.
Nvidia has a near monopoly in generative AI-capable GPUs. Cloud providers like AWS, Azure, and Google all use Nvidia’s H100 Tensor Core GPUs and add on services to help clients get projects using large language models up and running to differentiate themselves.
Microsoft and Nvidia also teamed up to build new supercomputers, even though Microsoft itself is reportedly looking to manufacture AI chips.
Nvidia also faces competition from AMD, which wants to increase production for its own AI GPU in the fourth quarter of this year.
Follow my posts on LinkedIN
