- Amazon Web Services (AWS) and Nvidia are expanding their partnership to significantly boost infrastructure capacity for generative AI, AI Agents, and physical AI.
- AWS plans to deploy an additional 2 million Nvidia GPUs globally during the 2027–2028 period.
- This new plan follows a previous commitment to deploy over 1 million chips starting in 2026, after customer demand far exceeded forecasts.
- AWS will integrate next-generation CPUs based on the Vera architecture to support increasingly complex AI workloads.
- The system will also feature network upgrades with NVLink Fusion technology and high-bandwidth memory to improve performance on large-scale computing clusters.
- Amazon’s data analytics services utilizing Nvidia’s cuDF software achieve processing speeds nearly 3.7 times faster than standard configurations.
- This upgrade simultaneously improves price-performance by about 30% compared to systems using only standard chips.
- Amazon’s vector search service reduces index-building time by approximately 9 times when running on GPUs.
- Amazon’s robotics division is collaborating with Nvidia to develop simulation tools to accelerate the training of next-generation warehouse robots.
- AWS has dedicated 100,000 chips to serve the sensitive computing needs of government and defense agencies.
- According to Nvidia, the new hardware systems achieve up to 4.6 times faster inference speeds and 2.1 times higher graphics performance than the previous generation.
- AWS and Nvidia leaders noted that accelerated computing demand continues to exceed forecasts, reflecting confidence that AI spending will rise sharply for years to come.
📌 Amazon and Nvidia are placing a massive bet on the long-term growth of AI. Adding 2 million GPUs on top of the previous 1-million-chip commitment, along with upgrades to CPUs, networking, and software, reflects an infrastructure expansion race at an unprecedented scale. With processing performance up to 3.7 times higher, inference 4.6 times faster, and 100,000 chips dedicated to the government, AWS aims to serve enterprises, AI labs, and the public sector in the 2027–2028 period.
