NVIDIA’s New Cloud Services to Make it Rain

NVIDIA isn’t clinging on to just one cloud partner - the hardware-software package is expected to be hosted by Microsoft Azure in the next quarter and then Google Cloud after that.
Listen to this story

Jensen Huang’s NVIDIA, which dominates the GPU market, had one noticeable trait that stood out at its annual GTC conference. For a company known for building AI hardware, it made a bunch of announcements related to cloud. NVIDIA is shifting hard and fast to a new business model where it plans to sell AI services through cloud computing platforms of companies such as Google and Microsoft Azure. The move has placed NVIDIA in a direct contest with these Big Tech companies that Huang’s company sells GPUs to. 

Even if NVIDIA isn’t naive enough to build an entire cloud infrastructure from scratch to rival hyperscalers like AWS, Microsoft Azure and Google Cloud, it is perceptive enough to realise how much cloud computing sells. In brief, NVIDIA’s DGX Cloud announcement wants to cash in on this compute. 

In competition with cloud providers

Just last week, a Bloomberg report stated that Microsoft had ended up spending hundreds of millions of dollars on tens of thousands of NVIDIA A100 GPUs for an OpenAI supercomputer. Now, as announced at the GPU Technology Conference, NVIDIA’s DGX Cloud wants to sell remote web access to users as a service. 

Starting at a price of USD 36,999 a month for the A100 tier, the service rents virtual versions of its DGX Server boxes, each of which contains eight NVIDIA H100 GPUs and 640GB of memory. The service can be scaled up to around 32,000 GPUs, storage, software and includes ‘direct access to NVIDIA’s AI experts who optimise your code.’ 

NVIDIA has partnered with Oracle Cloud and Equinix data centres aside from leveraging their own on-premises DGX SuperPod data centre platform. But NVIDIA isn’t clinging on to only one cloud partner—the hardware-software package is expected to be hosted by Microsoft Azure in the next quarter and then Google Cloud after that

The service is a way for NVIDIA to milk the generative AI fever that has taken over enterprises that either want to build or integrate generative AI tools within their workflow but hit a wall when it comes to scaling. 

NVIDIA has also jumped into another collaboration with Microsoft Azure for building their industrial metaverse network, ‘Omniverse Cloud’. The service will give users access to a full-stack environment to design, develop and manage industrial metaverse applications. Interestingly, only four months back, Satya Nadella-led Microsoft gave up on its own industrial metaverse project and laid off the entire unit. 

Catering to generative AI platforms

NVIDIA also launched ‘AI Foundations’, a new range of cloud services marked individually for different functions especially within the realm of generative AI—NVIDIA NeMo for LLMs, NVIDIA Picasso for images, videos and 3D content generation which also includes pretrained models, frameworks for data processing, APIs and engineering support from NVIDIA. Once these models are ready to be deployed, enterprises can just run them on NVIDIA’s cloud. 

This neat move could redirect a host of new generative AI startups to forgo the biggest cloud companies and go directly to NVIDIA. Huang is very optimistic about these new revenue streams and expects the money from the generative AI segment to grow from a ‘single digit’ portion of the company’s revenue to a ‘quite large’ portion of its revenue within the next year. While it’s also natural for the hyperscalers to look unfavourably upon this, NVIDIA has roped in every hyperscaling cloud platform as a part of the deal

NVIDIA’s sneaky entry into cloud

NVIDIA’s plans to diversify its services also come as some of their biggest buyers are designing their own chips to eventually manage the growing demands for data crunching in AI. For instance, Google designs its own AI chips now and it is highly likely that Microsoft will soon follow suit. This would inevitably lead to lesser demands for NVIDIA’s chips in the long term. 

Despite its wholly hardware origins, NVIDIA has been committed to a slow and steady march towards cloud. Given it sells GPUs that are very heavy on the pocket, NVIDIA was able to grab market share owing to its prescient decision to design and market its GPUs for cloud-based AI applications just as the demand for cloud services soared. 

In 2020, NVIDIA completed the acquisition of cloud computing company ‘Mellanox Technologies’ for a rather expensive price tag of USD 6.9 billion. It was merely a hint that NVIDIA was stepping foot outside of its GPU ring. 

It’s also a sign of how the rise of cloud computing has altered the scheme of things and how companies like NVIDIA that were able to look at the grand picture with fresh eyes were able to find new opportunities. 

Download our Mobile App

Poulomi Chatterjee
Poulomi is a Technology Journalist with Analytics India Magazine. Her fascination with tech and eagerness to dive into new areas led her to the dynamic world of AI and data analytics.

Subscribe to our newsletter

Join our editors every weekday evening as they steer you through the most significant news of the day.
Your newsletter subscriptions are subject to AIM Privacy Policy and Terms and Conditions.

Our Upcoming Events

15th June | Online

Building LLM powered applications using LangChain

17th June | Online

Mastering LangChain: A Hands-on Workshop for Building Generative AI Applications

Jun 23, 2023 | Bangalore

MachineCon 2023 India

26th June | Online

Accelerating inference for every workload with TensorRT

MachineCon 2023 USA

Jul 21, 2023 | New York

Cypher 2023

Oct 11-13, 2023 | Bangalore

3 Ways to Join our Community

Telegram group

Discover special offers, top stories, upcoming events, and more.

Discord Server

Stay Connected with a larger ecosystem of data science and ML Professionals

Subscribe to our Daily newsletter

Get our daily awesome stories & videos in your inbox

Is Sam Altman a Hypocrite? 

While on the one hand, Altman is advocating for the international community to build strong AI regulations, he is also worried when someone finally decides to regulate it