View and manage resources together
Manage physical servers, virtual machines, containers, Kubernetes and multi-cloud resources, view their status and handle routine management tasks.
Pool
BitPods
Enterprise cloud resource and inference service management
Manage resources together. Support enterprise inference services.
Bring distributed resources into a manageable service foundation.
How can physical servers, virtual machines, containers and cloud resources be managed together? How can existing GPUs support internal model services? BitPods brings resource management, inference services and daily operations together, helping teams manage resources and services on their existing infrastructure.
- Physical servers
- Virtual machines
- Containers & Kubernetes
- Multi-cloud resources
BitPods
Resource & service management
This illustrates the scope of unified management; resource types may require different onboarding processes.
From resource onboarding to running inference services.
Organize infrastructure in a unified management view, and manage inference services around models, resources and access endpoints.
Create inference services around models
Manage models in the model library, select engines, hosts and GPU resources when creating services, and view service status.
- Engine
- Host
- GPU
Manage access for business applications
Manage service access through user-level API keys and a compatible gateway, allowing applications to access inference instances through the corresponding interfaces.
Availability depends on the deployed version, resource environment, model and engine conditions.
Keep resource onboarding and daily maintenance organized.
Organize daily operations around node onboarding, host maintenance and state recovery.
Node onboarding
Connect compute nodes according to environment requirements, and organize existing GPU resources with batch onboarding tools.
Operations & maintenance
View resource and service status, and carry out host maintenance and related management tasks.
State & recovery
Perform recovery and remote power operations according to node status and management configuration.
Manage resources and services on your existing infrastructure.
Support enterprise cloud and inference teams with resource consolidation, service setup and daily operations.
Consolidate enterprise resources
Bring distributed hardware, virtualization and cloud resources into a unified view for daily management.
Build internal inference services
Manage models and services on existing GPU resources, and configure access for internal applications.
Manage compute nodes
Coordinate operations and delivery around node onboarding, host maintenance and state recovery.
Resource management and deployment validation, each with a focus.
BitPods
Resource and inference service management.
BitTune
Deployment, tuning and result validation.
FAQ
Which resources can be connected?
Physical servers, virtual machines, containers, Kubernetes and multi-cloud resources can be managed. Onboarding methods and availability depend on the hardware, cloud platform, network and deployed version.
Is Kubernetes required first?
Not always. The container-host approach can manage workloads through containerd and CNI; Kubernetes cluster management is a separate usage option. Choose the deployment approach for your workloads and existing environment.
Can existing GPUs run models directly?
First check the hardware, drivers, model format, inference engine and resource conditions. After GPUs are brought under management, service operation still needs validation against the model and workload; onboarding alone does not establish model compatibility.
Can inference services be accessed through an OpenAI-compatible interface?
The platform provides an OpenAI-compatible gateway and user-level API keys for inference instances. Check specific interfaces, parameters, streaming responses and model capabilities against the deployed version and engine.
How does it differ from Atlas, BitTune and BitCloud API?
BitPods manages resources and inference services. Atlas focuses on capacity assessment, selection and planning; BitTune on deployment, tuning and validation; BitCloud API on multi-model access and usage operations. Choose products for the task, and confirm combinations for the project.
Does it include unified Token billing?
This page does not list cross-system Token billing and settlement as delivered capabilities. Billing and integration requirements can be scoped during solution discussions.
How do we get started?
Share your existing hardware and cloud environment, planned models or workloads, and resource management and operations needs through the contact email to discuss onboarding and deployment scope.
