AWS joins up with Nvidia to build fastest AI supercomputer

The US companies announce at re:Invent in Las Vegas that they will work together on Project Ceiba

California-based Nvidia designs and manufacturers AI hardware and software GPUs for various industries. AFP

Amazon Web Services, the world's biggest cloud services provider, has joined forces with technology firm Nvidia to build the world’s fastest artificial intelligence supercomputer, the companies said.

The US companies will work together on Project Ceiba to design the GPU-powered AI supercomputer.

Hosted by AWS for Nvidia’s own research and development team, this supercomputer will feature 16,384 Nvidia GH200 superchips, Adam Selipsky, head of AWS, and Jensen Huang, chief executive of Nvidia, announced.

The new machine will be capable of processing 65 exaflops of AI – a measure of supercomputer performance – and will be used by Nvidia to propel its next wave of generative AI innovation.

“AWS and Nvidia have collaborated for more than 13 years, beginning with the world’s first GPU [graphics processing units] cloud instance,” Mr Selipsky said.

“Today, we offer the widest range of Nvidia GPU solutions for workloads including graphics, gaming, high performance computing, machine learning and now, generative AI.”

California-based Nvidia designs and manufacturers AI hardware and software GPUs for various industries. GPUs can process various tasks simultaneously, making them useful for machine learning, video editing and gaming applications.

The latest partnership, announced at AWS’s ongoing annual re:Invent event, also aims to build advanced infrastructure, software and services to facilitate generative AI innovations by other technology companies.

Currently, the development of generative AI technology is concentrated within a handful of large companies based in the US, the world’s largest economy.

Microsoft-backed OpenAI’s ChatGPT and Google’s Bard are two front-runners in the burgeoning field. In September, the Abu Dhabi government-supported research centre the Technology Innovation Institute launched Falcon 180B – an advanced version of its flagship language model – to boost generative AI capabilities in the region.

AWS and Nvidia will collaborate to host Nvidia DGX cloud, Nvidia’s AI-training-as-a-service model, on the AWS platform, as per the agreement. It will be the first DGX cloud featuring Nvidia’s latest multi-node platform that provides developers the largest shared memory in a single instance.

“Generative AI is transforming cloud workloads and putting accelerated computing at the foundation of diverse content generation,” said Mr Huang, who founded Nvidia in 1993.

“Driven by a common mission to deliver cost-effective, state-of-the-art generative AI to every customer, Nvidia and AWS are collaborating across the entire computing stack, spanning AI infrastructure, acceleration libraries, foundation models, to generative AI services,” he said.

Generative AI could add nearly $4.4 trillion annually to the global economy and will transform productivity across sectors with continued investment in the technology, according to a study by consultancy McKinsey.