From innovation to impact: How AWS and NVIDIA enable real-world generative AI success
As we gather for NVIDIA GTC, organizations of all sizes are at a pivotal moment in their AI journey. The question is no longer whether to adopt generative AI, but how to move from promising pilots to production-ready systems that deliver real business value. The organizations that figure this out first will have a significant competitive advantage—and we’re already seeing compelling examples of what’s possible.
Consider Hippocratic AI’s work to develop AI-powered clinical assistants to support healthcare teams as doctors, nurses, and other clinicians face unprecedented levels of burnout. During a recent hurricane in Florida, their system called 100,000 patients in a day to check on medications and provide preventative healthcare guidance–the kind of coordinated outreach that would be nearly impossible to achieve manually. They aren’t just building another chatbot; they are reimagining healthcare delivery at scale.
Production-ready AI like this requires more than just cutting-edge models or powerful GPUs. In my decade working with customers’ data journeys, I’ve seen that an organization’s most valuable asset is its domain-specific data and expertise. And now leading our data and AI go-to-market, I hear customers consistently emphasize what they need to transform their domain advantage into AI success: infrastructure and services they can trust—with performance, cost-efficiency, security, and flexibility—all delivered at scale. When the stakes are high, success requires not just cutting-edge technology, but the ability to operationalize it at scale—a challenge that AWS has consistently solved for customers. As the world’s most comprehensive and broadly adopted cloud, our partnership with NVIDIA’s pioneering accelerated computing platform for generative AI amplifies this capability. It’s inspiring to see how, together, we’re enabling customers across industries to confidently move AI into production.
In this post, I will share some of these customers’ remarkable journeys, offering practical insights for any organization looking to harness the power of generative AI.
Content creation represents one of the most visible and immediate applications of generative AI today. Adobe, a pioneer that has shaped creative workflows for over four decades, has moved with remarkable speed to integrate generative AI across its flagship products, helping millions of creators work in entirely new ways.
Adobe’s approach to generative AI infrastructure exemplifies what their VP of Generative AI, Alexandru Costin, calls an “AI superhighway”—a sophisticated technical foundation that enables rapid iteration of AI models and seamless integration into their creative applications. The success of their Firefly family of generative AI models, integrated across flagship products like Photoshop, demonstrates the power of this approach. For their AI training and inference workloads, Adobe uses NVIDIA GPU-accelerated Amazon Elastic Compute Cloud (Amazon EC2) P5en (NVIDIA H200 GPUs), P5 (NVIDIA H100 GPUs), P4de (NVIDIA A100 GPUs), and G5 (NVIDIA A10G GPUs) instances. They also use NVIDIA software such as NVIDIA TensorRT and NVIDIA Triton Inference Server for faster, scalable inference. Adobe needed maximum flexibility to build their AI infrastructure, and AWS provided the complete stack of services needed—from Amazon FSx for Lustre for high-performance storage, to Amazon Elastic Kubernetes Service (Amazon EKS) for container orchestration, to Elastic Fabric Adapter (EFA) for high-throughput networking—to create a production environment that could reliably serve millions of creative professionals.
Key takeaway
If you’re building and managing your own AI pipelines, Adobe’s success highlights a critical insight: although GPU-accelerated compute often gets the spotlight in AI infrastructure discussions, what’s equally important is the NVIDIA software stack along with the foundation of orchestration, storage, and networking services that enable production-ready AI. Their results speak for themselves—Adobe achieved a 20-fold scale-up in model training while maintaining the enterprise-grade performance and reliability their customers expect.
READ the latest news shaping the Nvidia market at Newsvidia
From innovation to impact: How AWS and NVIDIA enable real-world generative AI success, source
