The project, announced Monday (Nov. 24) will see Amazon Web Services (AWS) invest at least $50 billion as it works to build and deploy what it said is the first-ever artificial intelligence (AI) and high-performance computing (HPC) purpose-built infrastructure for the U.S. government.
This investment, Amazon said in a news release, is aimed at adding close to 1.3 gigawatts of compute capacity across AWS Top Secret, AWS Secret, and AWS GovCloud (US) Regions “across all classification levels.”
It also expands access to AWS’s infrastructure and AI services to help government agencies “advance America’s AI leadership,” the release added.
“Our investment in purpose-built government AI and cloud infrastructure will fundamentally transform how federal agencies leverage supercomputing,” said AWS CEO Matt Garman.
“We’re giving agencies expanded access to advanced AI capabilities that will enable them to accelerate critical missions from cybersecurity to drug discovery. This investment removes the technology barriers that have held government back and further positions America to lead in the AI era.”
Advertisement: Scroll to Continue
The release argued this investment underlines the importance of AI and supercomputing in upholding technological superiority, protecting critical infrastructure, and promoting industrial innovation.
“Federal customers and the supporting industrial base share a vision of AI and HPC convergence,” the company said. “This includes orchestrating expert AI models, agents, and natural language interfaces to enable researchers and engineers to explore complex problems through conversational interaction.”
That marks a shift from traditional HPC workflows to “AI-accelerated discovery,” in which scientists can specify challenges and get back AI-driven recommendations supported “by high-fidelity simulations and analysis,” the release continued.
In other Amazon AI news, PYMNTS wrote last week about the company’s Vulcan robot as an example of the way physical AI is moving from research to frontline operations.
“Earlier robots followed fixed commands and worked only in predictable environments, struggling with the unpredictability found in everyday operations such as shifting layouts, varying item shapes, mixed lighting, and human movement,” that report added.
“That is beginning to change as research groups show how simulation, digital twins and multimodal learning pipelines enable robots to learn adaptive behaviors and carry those behaviors into real facilities with minimal retraining.”
