While Big Tech giants like Google and Meta and heavily-funded AI startups like OpenAI and Anthropic have raced to grab GPUs, Nvidia has also focused on the advertising industry as one of its next big markets.
In May of this year, Nvidia’s Jamie Allan — who was previously responsible for leading partnerships within the EMEA media, entertainment, and broadcast industries — moved into a newly created role of director of business development for global agencies and advertising.
“When we see a particular industry or part of an industry on a significant growth curve with technology or AI or any form of accelerated computing, we look to assign resources and focus into that area,” Allan said.
When working on specific advertising accounts, Allan will typically spin up a virtual team of around four staffers to build products using Nvidia’s software and hardware or to match them with technologies developed by its partners. Microsoft’s Bing Ads used Nvidia’s software to speed up the delivery of personalized ads, for example. It was able to increase the amount of information it could process within a given amount of time seven times over, and at a reduced cost as of January, Allan said.
“Wherever and however they want to consume these technologies, we want to be supportive to move the industry forward with accelerated computing,” Allan said.
For Nvidia, the advertising industry unlocks access to a client base that includes virtually every single major corporation across every industry.
For instance, ad giant WPP had been working with Nvidia to build 3D applications, and extended that partnership to add generative AI capabilities in May. This allowed WPP to build a 3D car configurator for the Chinese automaker DENZA and generate thousands of individual pieces of content using generative AI, set in digitized environments.
“The combined 3D plus generative AI work we’re doing cannot be done on run-of-the-mill machines or publicly available cloud infrastructure,” said WPP CTO Stephan Pretorius, who said the configurators also help automakers create demand and fill order books earlier. He described Nvidia’s technology as “cutting-edge and massively scaled infrastructure that cannot be accessed without a deep and long-term partnership.”
Smaller advertising companies are also tightening their relationships with Nvidia.
“There’s definitely been an acceleration of working together with Nvidia,” said Adam Singolda, CEO of the adtech company Taboola.
Taboola is using Nvidia GPUs to speed up its natural language processing, which analyzes articles and serves relevant content to users. More recently, it’s been using Nvidia and OpenAI technologies to allow advertisers to create images within its platform.
Elsewhere, S4 Capital’s Media.Monks announced this month an AI software offering that harnesses Nvidia technology to identify and create custom highlights from live broadcasts to be served to different audiences.
Experts have said that more ad agencies are looking to shore up direct access to the GPU supply chain to capitalize on AI’s promise of delivering more personalized and more eye-popping ads. Some advertising companies are also taking steps to rent GPU capacity months in advance to ensure they’re not at the back of the queue in the face of shortages.
“Compute power is starting to be raised as an issue,” at advertising companies, said Jay Pattisall, vice president and principal analyst at the research agency Forrester. “If you look at the top 10 advertisers globally, it’s millions and billions of impressions, and each of them demanding some sort of asset or product from compute power — cloud, chip, or otherwise.”
To be sure, currently many advertising agencies and adtech companies are simply renting compute power via their cloud service providers like Amazon’s AWS or Oracle and haven’t yet sought to forge close relationships with chipmakers with the likes of Nvidia. Right now there are some pricing advantages to be had as cloud computing companies duke it out for AI market share.
Allan said Nvidia intends to be more involved in the advertising industry discourse and help advise companies about the adoption of AI. Nvidia CEO Jensen Huang spoke on stage at the advertising industry’s tentpole event Cannes Lions this summer, for example, and the company plans to return in 2024.
Next year, Allan predicts more advertising companies will shift from producing pilot projects using AI and large language models to embedding these technologies in their day-to-day services.
“The client base, whether it’s in retail, automotive, pharma, they’re expecting this now they’ve seen the power and they know what the capabilities can be,” Allan said.
Read the full article here