Campbell, Calif. – October 07, 2026 -- WEKA Federal, the wholly-owned subsidiary of WEKA, has been designated an "Awardable" vendor in the U.S. Department of War's Tradewinds Solutions Marketplace, clearing the way for defense acquisition teams to procure and deploy its NeuralMesh software without repeat technical evaluations.
Designation lets DoW program offices skip competitive re-evaluation
Tradewinds, housed within the DoW's Chief Digital Artificial Intelligence Office (CDAO), is a digital repository of post-competition, pre-vetted pitch videos addressing the department's AI/ML, data and analytics challenges. The Awardable status means mission owners can now award NeuralMesh contracts directly, bypassing a lengthy procurement cycle.
NeuralMesh promises up to 8x smaller infrastructure footprint
WEKA Federal said NeuralMesh can cut infrastructure footprint by up to 8x and power draw by up to 6x versus scale-out NAS architectures at AI-factory scale, freeing rack space, cooling capacity and megawatts for DoW missions. The platform also claims to reduce model training time by up to 20x and, combined with its Augmented Memory Grid technology, support up to 10x more concurrent users per GPU while lowering cost per token.
Single namespace spans accredited data centers, cloud and air-gapped sites
Backed by more than 170 patents, NeuralMesh runs on standard hardware and major public clouds, serving POSIX, S3, NFS, SMB and GPUDirect Storage protocols from one namespace scalable to exabytes. The design lets DoW teams move AI training, inference and analytics applications across accredited data centers, cloud regions and air-gapped environments without rewriting applications or maintaining separate infrastructure stacks. A native S3 object store shares the same NVMe blocks and namespace as file data, eliminating separate storage tiers, gateways or duplicate copies that require additional security management.
Executive frames deal as competition for compute capacity
"Every AI system the DoW fields is competing for the same megawatts, the same rack space, and the same accelerators,