Infrastructure as a Service, Engineered for Critical Workloads
Services
Compute
Virtual machines running on N+1 redundant hyperconverged nodes. Choose shared tenancy on the vetted critical-infrastructure platform, or dedicated clusters with no CPU or memory sharing.
GPU-equipped nodes are available for vendor AI/ML and numerical analysis workloads.
Storage
All-NVMe persistent block storage attached to virtual machines, and Kubernetes persistent volumes through a CSI driver for containerized platforms. High performance storage supports demanding applications such as time-series historians and relational databases.
Virtual Networking
Isolated virtual private clouds, administered through the ACS Console, with vendor-controlled subnets, routing, and security policy inside each tenant. The full network path from workload to internet edge or carrier handoff is documented and available to customers.
Connectivity
Encrypted IPsec VPN over the internet is the standard starting point for connecting utilities and field networks. For utilities that require off-internet connectivity, ACS facilities terminate private circuits (MPLS, dedicated fiber, carrier Ethernet) in dedicated meet-me rooms with defined physical demarcation. LEO satellite links are available to provide terrestrial connectivity independence.
Internet transit is multi-carrier on an ACS-owned autonomous system. ACS sites are joined by diverse private metro Ethernet carriers with single-digit millisecond latency.
Resilience and Disaster Recovery
Each site is free of hardware single points of failure. Vendors can deploy within one site, in active-passive across multiple sites, or active-active across multiple sites for high availability. Scheduled cross-site DR tests can be witnessed by the vendor or its utility customer.
Time
GPS-sourced stratum 1 NTP with holdover oscillators at each site, available to every workload, with a published accuracy commitment.
Monitoring and Security Operations
ACS automated monitoring scans thousands of health points 24×7 and alerts on issues within seconds. ACS staff are alerted through multiple channels, including off-internet pathways, depending on event severity. Security event monitoring is staffed 24×7 by a specialized managed security operations center. Tenants receive incident notifications specific to their environment with OT-relevant impact context.
Facilities
Virginia
A dedicated, access-controlled data room within a hardened, carrier-neutral colocation facility built for US federal government customers. Biometric entry, 24×7 monitored video, and physical controls mapped to NERC CIP-006. Multi-carrier transit and private circuit termination on site. 120+ hours of on-site generator fuel.
Maryland
A data facility built and operated by ACS, giving ACS control of the physical design from the walls in. Independent utility and carrier service from the Virginia site. 120+ hours of on-site generator fuel; N+1 power and cooling.
ACS sites are served by different transmission system operators and different carrier hotels, so a regional grid or carrier event is unlikely to affect more than one.
Pricing Model
Compute and storage are billed as fixed monthly capacity, sized to the vendor’s requirements. Network transit is billed on a 95th-percentile basis at published rates. There is no metering of API calls, storage operations, or CPU overages, so the monthly invoice can be modeled in advance and stays stable month to month.
This suits critical applications, whose load is consistent and whose resources must be guaranteed rather than allocated when available.
Want the Architecture Detail?
Physical and logical architecture diagrams with security boundary demarcation, documented network paths, and the hardware vendors are available under NDA.