National Compute Grid has launched a platform that pools idle AI compute capacity across multiple GPU providers.

For all the talk about a shortage of AI computing power and the need for more powerful chips and bigger data centers, there is another side to the story. A significant amount of existing capacity goes unused. While tech companies race to build massive new data centers, expensive GPUs in other facilities can sit idle between workloads.

What if all that fragmented computing capacity could be pooled to become something much bigger? That is the goal of an ambitious new initiative called the National Compute Grid. It is a nationwide network of companies capable of pooling GPU resources and making them available wherever they are needed, at least in concept.

The idea is that instead of always building more data centers or buying more powerful chips, the industry could make better use of the computing resources already available.

Here is how it would work. Data centers and cloud providers can make their unused computing capacity available through a software platform called Grid Exchange. The network is open to AI companies, startups, researchers, and other users who need GPUs. The Grid matches them to available computing resources based on the type of hardware required, available capacity, and what customers are willing to pay.

The grid is designed to support different chip architectures, including GPUs from NVIDIA and AMD and TPUs from Google. That gives customers more options. However, moving workloads between different types of hardware is not always straightforward because of software compatibility issues.

The project is already substantial, at least on paper. According to its organizers, the grid has around 760 megawatts of computing capacity either connected or in the pipeline. The goal is to reach 2 gigawatts by 2030.

The organizers also say that single-tenant data centers use less than 15% of their computing capacity on average. That does not mean 85% of all GPUs are sitting idle, but it does suggest how much computing power may go unused at some facilities.

The grid could be particularly useful for smaller AI companies. Unlike OpenAI and Anthropic, which can afford to sign large, long-term computing contracts, smaller companies do not always have that luxury. Sam Sinha, head of AI at robotics company 1X, discussed this problem with Axios. For a smaller developer that needs extra GPUs for a few weeks or a particular project, having somewhere else to turn could be a big help.

Government agencies, academic institutions, and national labs could also use the network. Some research projects require more computing power than an institution has available, and getting time on existing systems is not always easy. The grid may provide much-needed capacity for researchers.

It could also serve President Trump's Genesis Mission, which requires substantial computing power for AI research. National Compute plans to contribute $100 million in computing credits to the mission, giving researchers another way to access the GPUs they need.

Participating researchers can tap into computing resources across the grid, which reduces their dependence on government supercomputers or dedicated cloud contracts. It also gives National Compute an opportunity to show that its model can work for scientific workloads, not just commercial AI applications.

The timing is notable. The initiative was announced as the White House prepares to unveil more than $1 billion in private-sector commitments to the Genesis Mission. Companies including AMD, OpenAI, and Anthropic are expected to take part.

The initiative is being spearheaded by Anjney Midha, who leads the AI holding company Amp. Cloud infrastructure provider Vultr is a confirmed founding member of the consortium and is helping bring computing capacity to the network. National Compute Labs is developing the grid and its exchange platform. Other AI companies, infrastructure operators, and investors are involved, although a complete list of members has not been made public.

The consortium has taken a different approach to pricing. According to Vultr, the grid will support a model called "goodput," under which customers are charged for completed, uninterrupted computing work rather than for the number of GPU-hours used. National Compute has reportedly attracted more than $5 billion in reservations so far, but those reservations should not be confused with revenue already earned.

There will be challenges ahead, which is to be expected. Bringing together GPUs from different providers is one thing. Ensuring they can reliably handle demanding AI workloads is another challenge entirely.

Training large AI models often requires thousands of GPUs connected through extremely fast networks. Spare GPUs spread across multiple data centers do not necessarily work together for such jobs. Moving large datasets between facilities can also be expensive and time-consuming.

The grid may be most useful initially for smaller training runs, inference, and research workloads that do not require thousands of GPUs working in concert. Whether it can reliably support larger jobs remains to be seen.