Microsoft has released information about its latest virtual machine (VM) designed for artificial intelligence workloads, as the company continues to develop and release AI products following a significant investment in OpenAI, the company behind ChatGPT.
In a blog post, Microsoft emphasised the critical role its GPU-accelerated VMs play in laying the groundwork for many generative AI advancements, both in-house and from other companies, including its own Azure OpenAI Service, which puts AI writers such as GPT-3.5, Codex, and DALL-E 2 in the hands of its customers.
Microsoft promises “significantly faster performance” with its new ND H100 v5 virtual machine when compared to previous-generation ND A100 v4 virtual machines.
The updated VM makes use of eight NVIDIA H100 Tensor Core GPUs, with the ability to scale up to thousands. It also houses 4th Generation Intel Xeon Scalable processors and 16 channels of DDR5 DIMMs at 4800 MHz.
With this move, Microsoft hopes to bring some of the most cutting-edge, high-performance technology to AI startups, many of which will be comparing their own success stories to OpenAI’s recent success.
As the world strives to make artificial intelligence as reliable and accurate as possible, the ND H100 v5 VMs will need to be able to handle large amounts of training for extended periods of time. The new H100-powered VMs are now available for testing as the tech giant prepares to make them a standard offering in the Azure portfolio.