NVIDIA Technologies Transform Advertising at Cannes Lions 2023
During the Cannes Lions festival held from June 22-26, 2023, industry leaders showcased how NVIDIA technologies are revolutionizing advertising and marketing. Companies like Alembic, Amazon Web Services (AWS), Criteo, Higgsfield, KERV.ai, and Taboola demonstrated how AI is enabling faster and more autonomous operations across the advertising landscape. This shift is not merely about adopting AI; it’s about ensuring that infrastructure can handle the demands for speed and scale in a rapidly evolving market.
Decision Intelligence at Enterprise Scale
Alembic, a causal AI platform, addresses a significant challenge for enterprises: identifying which marketing initiatives genuinely drive growth rather than merely reporting on past performance. By employing advanced AI infrastructure capable of processing vast and dynamic datasets, Alembic models true causation across various channels, markets, and audiences.
The integration of NVIDIA DGX Vera Rubin NVL72 systems allows Alembic to enhance its Causal AI models by analyzing more variables and conducting larger simulations. This capability provides executives with a reliable source of truth regarding business outcomes and capital allocation. Alembic stands out as the first Causal AI company to utilize NVIDIA DGX Vera Rubin SuperPODs for large-scale causal modeling, empowering decision-makers with actionable insights.
Furthermore, Alembic’s inference operations occur within private supercomputing environments at Equinix data centers, ensuring that enterprise data remains secure and local. Collaborating with World Wide Technology enables these solutions to extend into secure environments, offering a comprehensive AI stack tailored for executives responsible for financial decisions.
Smarter Bidding at Auction Speed
For advertisers navigating billions of daily transactions, the need for accurate and rapid AI capabilities is paramount. AWS is addressing this demand by integrating cloud infrastructure with NVIDIA GPU-accelerated computing to create a robust framework for the adtech industry. This stack allows advertisers to implement AI-powered bidding directly within auction environments using NVIDIA Triton Inference Server, which ensures deep learning inference occurs swiftly enough to meet real-time auction requirements.
This advancement facilitates a transition from traditional rules-based decision-making to dynamic AI-driven models that optimize bid prices, activate audiences, and score deals in real-time auctions. Criteo exemplifies this shift by utilizing its extensive recommendation network to match products with shoppers precisely when they are most likely to convert. By collaborating with NVIDIA, Criteo has achieved nearly double the speed in model training on NVIDIA Blackwell GPUs through the use of the open-source NVIDIA cuEmbed library. This efficiency translates into approximately 17,000 GPU hours saved annually.
Additionally, Taboola is leveraging similar infrastructure strategies for conversational AI applications. By utilizing NVIDIA GPUs to power its DeeperDive engine, Taboola enhances its ability to generate revenue through advertising across various platforms.
Agentic AI Across the Marketing Workflow
The deployment of AI agents as digital coworkers is becoming more prevalent in marketing and other sectors. These agents streamline tasks throughout planning, execution, and optimization phases but require robust controls such as safety measures and role-based permissions to ensure responsible use.
NVIDIA’s Agent Toolkit provides essential controls through tools like NVIDIA NemoClaw blueprints and OpenShell secure runtime environments. Higgsfield AI exemplifies this approach by offering Supercomputer agents that manage the entire marketing automation lifecycle—from campaign ideation to performance analysis—within a single interface. These agents integrate leading large language models alongside various multimedia models built on NVIDIA Blackwell architecture.
The collaboration between Higgsfield and NVIDIA ensures that specialized subagents operate continuously within campaigns while maintaining enterprise-level trust through integrated security measures. As a result, marketing campaigns for nearly 400 Fortune 500 companies are efficiently created on this platform.
Contextual and Content Intelligence at Scale
Understanding content at an advanced level requires sophisticated infrastructure capable of processing complex data sets. NVIDIA’s multimodal stack offers the necessary vector search capabilities alongside data processing tools for video understanding at scale.
KERV.ai has optimized its Moment Match Engine to evaluate numerous signals across video frames and media assets effectively. This technology enables KERV.ai to provide content recommendations based on ad creatives—enhancing engagement by targeting specific moments within content environments accurately.
The recent optimization of KERV.ai’s processing pipeline resulted in over tenfold improvements in speed when utilizing the NVIDIA Nemotron 3 Nano Omni open model. This enhancement allows KERV.ai’s solution to analyze advertisement contents efficiently while identifying their resonance with target audiences.
What This Means
The advancements showcased at Cannes Lions 2023 highlight a critical transformation in how advertising and marketing industries leverage AI technologies. As companies increasingly adopt these innovations powered by NVIDIA technologies, they can expect enhanced decision-making capabilities driven by data-driven insights. The shift towards autonomous operations not only improves efficiency but also empowers marketers to make informed choices that align closely with consumer behavior—ultimately driving growth in an increasingly competitive landscape.
For more information, read the original report here.
