Google has upgraded its Google Cloud AI Hypercomputer, a platform offering advanced AI training and serving capabilities to its customers. The update includes more powerful Nvidia chips and a range of new software and storage solutions.
The AI Hypercomputer enhancements feature the general availability of Google Cloud's Tensor Processing Unit (TPU) v5p, the introduction of A3 Mega virtual machines powered by Nvidia H100 Tensor Core GPUs, and plans to incorporate Nvidia’s Blackwell GPUs. Key software updates include JetStream, an inference engine, and Hyperdisk ML, a block storage service optimized for AI.
Google claims these upgrades will significantly boost efficiency and productivity for AI model training, tuning, and serving, highlighting improved performance per dollar, reduced model load times, and higher throughput.
By using this site, you agree to allow SPEEDA Edge and our partners to use cookies for analytics and personalization. Visit our privacy policy for more information about our data collection practices.