Two companies that have been quietly building credibility with the US government just made their most ambitious joint move yet. Parallel Works, a spinoff from Argonne National Laboratory, and CoreWeave, the Nasdaq-listed AI cloud provider trading under CRWV, are combining forces to deliver a fully managed AI cloud environment purpose-built for DARPA’s biological research programs.
The collaboration pairs Parallel Works’ ACTIVATE platform with CoreWeave’s GPU-heavy cloud infrastructure to handle the kind of computationally brutal tasks that biological defense research demands. Think genomics sequencing, molecular modeling, and the sort of AI workloads that make a standard cloud setup weep quietly in the corner.
What the partnership actually does
Parallel Works’ ACTIVATE platform provides unified access and management across on-premises and cloud resources. The platform has a high-security variant specifically designed for defense applications, which matters enormously when you’re dealing with controlled unclassified information.
CoreWeave brings the raw computational muscle. Its AI cloud infrastructure features bare-metal NVIDIA GPUs and Kubernetes-native orchestration, giving researchers direct access to the kind of hardware that AI training and inference workloads actually need, without the virtualization overhead that general-purpose cloud providers typically impose.
Why these two companies, and why now
Parallel Works secured its IL5 Authorization to Operate, which is the security clearance needed to handle Controlled Unclassified Information and ITAR workloads. The company also holds contracts with the Department of Defense’s High Performance Computing Modernization Program and has worked with the Naval Research Laboratory on weather forecasting initiatives.
CoreWeave, meanwhile, has been making its own federal push. The company launched CoreWeave Federal as a dedicated initiative to expand its government engagements, and it’s been involved with the Department of Energy’s Genesis Mission. It also announced a collaboration with Leidos, one of the largest US defense contractors, further cementing its position in the government technology ecosystem.
The bigger picture for AI infrastructure in defense
Parallel Works occupies a specific niche. Its value isn’t in the hardware but in the orchestration layer, the software that makes it possible to run complex scientific workflows across heterogeneous computing environments without requiring researchers to become DevOps engineers.
The combination of the two, specialized AI hardware plus intelligent orchestration software, is designed to address a specific pain point that defense research labs face. They need supercomputer-class performance with cloud-native flexibility, wrapped in a security posture that satisfies government requirements.