NetThunder portfolio

Choose the right platform.

Start with where the work runs and what you need to manage. Spark™, SecureEnclave™, CloudSpawn™, and Storm™ are four distinct platforms for different operating needs.

Compare platforms

Workload, boundary, and capacity

Find your starting point.

Compare the operating shapes, then follow the platform that fits your work.

Spark™A complete environment on one supported node for operational edge sites and qualified private virtualization.One supported node SecureEnclave™A controlled environment connecting selected CUI workflows, safeguards, technical evidence, and recovery.A selected CUI boundary CloudSpawn™Elastic, supercomputer-class infrastructure orchestrated within a customer-controlled tenant environment.An elastic tenant environment Storm™Replace manual cluster configuration while retaining compatible hardware, then manage growth and software changes from one controller.A managed bare-metal cluster

Operational edge and private virtualization

Spark™

Spark™ places a complete, customer-controlled operating environment on one supported node for operational edge deployments, OT trust zones, and private virtualization workloads.

Operating shapeOne node connects every path.

OT networks
Devices
Workloads
Data paths

One supported nodeComplete environment

Private virtualizationApplications + servicesRequired + observed state

The selected design defines the trust zone, runtime, connections, capacity, and operating authority.

Best fit: Operational edge deployments, OT zones of trust, and qualified private-virtualization targets.
Operating shape: One supported node carrying one complete, customer-controlled operating environment close to the work.
What it manages: Configures the supported applications, services, resources, and dependencies on the node, then applies and records changes to their state.

Configuration details

Control and boundary: The design defines which networks, devices, workloads, data, administrative paths, and operating responsibilities belong together.

Hardware and storage
Hypervisor and runtime versions
Device and network support
Capacity and resilience
Connectivity and support paths
Migration and acceptance plan

The deployment design qualifies the selected host, runtime, paths, capacity, resilience, access, support, and validation criteria.

Explore the full Spark™ platform

CMMC and controlled workflows

SecureEnclave™

SecureEnclave™ makes the selected CUI boundary operable as one controlled environment across applications, safeguards, identity, telemetry, recovery, and responsibility.

Operating shapeWorkflows define the controlled boundary.

Approved pathIdentity + policy

Selected CUI boundary

Approved applicationsProtected dataManaged stateTelemetry + evidenceRecovery scopeOperating controls

Human responsibility Authorization · risk · exceptionsIndependent assessment In-scope implementation

The approved design keeps technical state, evidence, recovery, and responsibility connected without turning human judgment into an automated claim.

Best fit: CUI and sensitive workflows whose applications, safeguards, evidence, recovery, and responsibilities must stay connected.
Operating shape: One controlled environment and managed boundary organized around the selected CUI workflow.
What it manages: Maintains the selected environment's configuration, safeguards, technical evidence, and supported recovery operations as its state changes.

Configuration details

Control and boundary: Physical, logical, identity, administrative, data, evidence, recovery, and responsibility scope are defined together while authorization and assessment remain human decisions.

In-scope people, systems, and data
Approved applications and safeguards
Identity and administrative paths
Connectivity and transfer modes
Telemetry and technical evidence
Recovery tests and responsibilities

The approved design defines the CUI scope, components, safeguards, connectivity, evidence, recovery, and operating responsibilities.

Explore the full SecureEnclave™ platform

Elastic supercomputing orchestration

CloudSpawn™

CloudSpawn™ orchestrates elastic, supercomputer-class infrastructure as a complete software-defined tenant environment across supported bare-metal compute, accelerators, high-performance data fabrics, and parallel storage.

Cloud-provider physical service remains outside tenant administration. Scoped support follows the procedures approved for the tenant environment.

Best fit: AI, research, engineering, and HPC workloads that need workload-shaped bare-metal compute, accelerator capacity, and high-performance data paths with hyperscaler-style elasticity.
Operating shape: An isolated, software-defined tenant environment serving elastic, supercomputer-class bare-metal infrastructure as workload requirements change.
What it manages: Orchestrates tenant resources, services, identity, policy, and lifecycle as workload and capacity requirements change.

Configuration details

Control and boundary: Isolation spans resources, identity, state, data, control paths, telemetry, support, and physical service, separating the tenant environment from other tenants and cloud-provider administration.

Workload, data, and services
Capacity profile and topology
Supported bare-metal resources and fabrics
Parallel storage and workload services
Lifecycle approvals and failure handling
Recovery and acceptance criteria

The tenant design qualifies supported resource classes, topology, services, isolation, lifecycle controls, support procedures, and validation criteria.

Explore the full CloudSpawn™ platform

Cluster management

Storm™

Storm™ is NetThunder's scheduler-agnostic cluster-management software for provisioning supported bare-metal infrastructure and maintaining configuration as nodes, applications, and requirements change.

The selected design defines the supported resources, application scope, out-of-band control path, lifecycle actions, performance criteria, responsibilities, and acceptance tests.

Best fit: Existing and new HPC, research, engineering, and AI clusters whose owners want repeatable deployment and less manual configuration work.
Operating shape: Controller software, supplied on physical controllers or as installation images for qualified customer nodes, that configures and manages a bare-metal cluster.
What it manages: Re-images selected bare-metal nodes, configures storage-backed boot and user-defined software stacks, and calculates changes from software packages and their dependencies.

Configuration details

Control and boundary: The cluster specification defines resources, services, control paths, approvals, state, and operating responsibilities. Any enclosing tenant or security boundary remains explicit.

Resource classes and topology
High-performance fabric and storage paths
Services and dependencies
User-defined schedulers and software stacks
Lifecycle approvals
Failure and recovery behavior

The deployment design qualifies the cluster resources, integrations, lifecycle actions, control paths, responsibilities, and acceptance criteria.

Explore the full Storm™ platform

An integrated AI appliance

Choose where AI runs.

Storm AI Factory™ is an AI operations module offered as a standalone appliance for your existing infrastructure, on premises or in colocation. It includes Storm™ infrastructure management and open-source AI software for inference, training, and agents.

It is also included with CloudSpawn™ and available in selected Spark™ and SecureEnclave™ configurations. Storm™ cluster owners can run AI applications without the AI Factory appliance.

Storm AI Factory™

Run AI workflows.

Create controlled sandboxes, train models, run inference, and operate agents within the selected environment.

Spark™: Selected configurations
SecureEnclave™: Selected configurations
CloudSpawn™: Included

Review the AI Factory module

Operating authority

Know who controls what.

A customer-controlled operating environment defines operating authority and responsibilities independently from hardware title.

Location: Where infrastructure is placed and where data is stored or processed.
Administration: Who administers the environment and which privileged paths are permitted.
Identity and policy: Who may access the environment, which connections are allowed, and how changes are approved.
Provider access: What access a provider requires, when it may be used, and how it is recorded.
Observation and audit: What the customer can observe, approve, restrict, isolate, and audit.
Title and financing: Record commercial and hardware-title terms as their own control dimension alongside operating authority.

Architecture review

Match platform to workload.

Tell us what needs to run and where. Start with a conversation about platform fit, open questions, and the design decisions ahead.

Request an architecture review Use the planning guide

Email NetThunder +1 847.477.7676