One cluster per customer
Flatkey Inc. has launched Flatkey Compute, a service that builds and operates NVIDIA B300 clusters, each reserved for a single customer and shared with no one, the company said. Flatkey runs every cluster as a compliant inference cloud for AI workloads and does not offer general-purpose computing.
The company said the arrangement removes co-tenants from the picture, so another customer's workload cannot affect a cluster's performance and a security review has no other tenants to account for.
Sizing and operations
GPU count, region and network topology are set per deployment rather than selected from a list of standard instance types, according to the release, and each cluster is sized to the customer's workload. Once a cluster is live, Flatkey operates it, so the customer does not need its own infrastructure team to run it. Ahead of the launch, Flatkey said it had delivered compute capacity to dozens of major customers.
On data protection, Flatkey said it is GDPR compliant, monitors its security controls continuously through Vanta and does not retain request content.
Terms and availability
Customers request capacity at flatkey.ai/compute, stating the GPU count, region and term they need. Flatkey said it replies within 24 hours and then sends a formal quote setting the price, configuration, region and provisioning date. Reserved terms of one, three and five years are available, with longer commitments priced lower, and a reference GPU price index on the same page shows indicative rates before a customer gets in touch.
The service is available now by request. Smaller teams that need only a few GPUs can rent them on demand, billed by the hour, in the Flatkey console at console.flatkey.ai/gpu-cluster.
