Amazon Web Services CEO Matt Garman recently sat down with CNBC’s Jon Fortt to discuss the company’s latest revenue growth, its strategic positioning in the burgeoning artificial intelligence landscape, and the challenges posed by AI supply constraints. The conversation offered a candid look into AWS’s approach to the current technological inflection point and its long-term vision for cloud infrastructure.
Garman articulated a powerful vision for the future of AI, emphasizing its profound impact across industries. He stated that AI is "such a transformative technology that is really going to change everything that customers do." AWS is actively "laying this groundwork" for what it believes will be the true value for enterprises from AI, focusing on modernization and cloud adoption over the next several years, particularly through "agentic workflows" that drive measurable ROI. This reflects a strategic pivot toward enabling complex, intelligent automation for its vast customer base, underpinning the next wave of enterprise innovation.
However, the rapid acceleration of AI adoption has presented its own set of challenges, particularly concerning the supply of critical hardware. Garman openly admitted that AWS faces supply constraints in certain areas, especially for customers building "very large training clusters" that demand "massive amounts of compute." He candidly acknowledged, "No one can say yes to everyone."
This bottleneck is an industry-wide phenomenon, driven by unprecedented demand for specialized AI chips and the underlying infrastructure. Garman explained that while AWS strives to support all customers, it prioritizes "mission-critical" enterprise and startup workloads, ensuring continuity for core business operations. He noted that constraints can stem from various components—chips, power, or other supply chain elements—and that AWS is relentlessly working to mitigate these.
Despite the supply pressures, AWS maintains a robust partnership with Nvidia, a key supplier of AI GPUs. Garman clarified that Nvidia’s DGX Cloud, a high-performance computing solution, "runs in AWS," and that Amazon is "one of their biggest DGX Cloud partners." He further highlighted that AWS adds significant value to Nvidia’s offerings through its proprietary "Nitro" system, which enhances enterprise security, isolation, and encryption for EC2 instances. This collaborative yet distinct approach underscores AWS’s commitment to providing differentiated, secure, and high-performance AI infrastructure, whether leveraging third-party hardware or its own custom silicon.
