Amazon Validates Custom Silicon Strategy with Massive Meta and Anthropic AI Deals

Amazon (AMZN) is cementing its position in the artificial intelligence infrastructure race by leveraging its custom silicon strategy to attract industry titans. The company recently announced a major partnership with Meta Platforms, which will deploy tens of millions of cores of AWS Graviton5 CPUs to power its next-generation AI workloads. While the AI narrative has been dominated by GPUs for model training, the industry is shifting toward inference and agentic workloads where CPUs play a critical role. The Meta deal serves as a high-profile validation of Amazon's ability to provide high-performance, cost-effective alternatives to traditional hardware providers. The momentum extends to Anthropic, which has committed over $100 billion to AWS technologies over the next decade. This agreement includes the use of Graviton and Trainium2 through Trainium4 chips, supported by 5 gigawatts of new capacity. Amazon is further deepening this tie with an immediate $5 billion investment in Anthropic, with a potential for an additional $20 billion in the future. Financial metrics highlight the scale of this growth. CEO Andy Jassy reported that the AWS AI revenue run rate exceeded $15 billion in the first quarter of 2026. More impressively, the broader chip business—comprising Graviton, Trainium, and Nitro—has reached an annual revenue run rate of over $20 billion, exhibiting triple-digit year-over-year growth. These developments have fueled a rally in Amazon's shares, which have climbed more than 25% over the last 30 days. Investors are increasingly confident that the company's aggressive capital expenditure in AI is yielding tangible returns through diversified hardware demand and strategic enterprise partnerships.