How Dell EMC and NVIDIA aim to simplify the AI data centre
FYI, this story is more than a year old
Many organisations have begun to infuse AI into their products, services and supply chains, but IT departments need to scale these efforts in a systematic way.
The newest DGX reference architecture, the Dell EMC Isilon with NVIDIA DGX-1, is an integrated infrastructure offering that supposedly does just that.
Sold through a joint channel, it supposedly gives customers a new approach for rapidly deploying AI at scale in the data centre.
The dramatic rise in AI compute deployments, often led by line-of-business leaders and accomplished within silos, has led to AI infrastructure sprawl.
It’s much like the early days of cloud computing, where a consolidated, optimised, shared infrastructure was lacking, forcing teams to seek their own solutions.
Businesses are realising they need AI at scale, and so enterprise IT teams are increasingly inserting themselves into their company’s AI agenda.
To scale efficiently in the data centre, like any other business-critical application needs to do, they’re using NVIDIA DGX-1 as their AI compute standard.
NVIDIA DGX systems supposedly give enterprises the fastest start in AI development and performance powered by an eight-way NVLink configuration that offers increased performance and fastest time to solution.
NVIDIA DGX POD is an IT-standardised approach to infrastructure that enables data science teams to train bigger models and datasets that incorporate more features.
Using NVIDIA DGX POD, IT leaders and their line-of-business leaders can supposedly bring important benefits to their organisations:
These benefits have a democratising effect on an organisation’s ability to spread machine learning and deep learning to the furthest reaches of the enterprise. And this new solution from Dell EMC and NVIDIA help them get there.
Built on the experience and best practices implemented in DGX POD, this powerful combination of Dell EMC Isilon storage and NVIDIA DGX-1 compute, data centre IT now has a proven solution that supposedly gets AI deployed faster. It offers predictable performance that scales and fits enterprise within their operations.