Skip to content

d.run Token Factory

d.run Token Factory is an integrated platform built by DaoCloud for AI intelligent computing scenarios, providing end-to-end capabilities around large model inference, compute scheduling, and cluster operations. The platform adopts a dual-view architecture of User and Admin: the user view serves end users, providing large model service invocation, cost management, and intelligent application access capabilities; the admin view serves operations and management teams, providing unified management of model hosting, inference engines, infrastructure, and global operations.

CNAI
ClawOS Multi-agent runtime and governance AI Apps Intelligent Q&A and other application capabilities LLM Studio Model serving and O&M management InferX Inference Inference acceleration and engine management
redhare Distributed Cache Distributed cache acceleration service
Compute Cloud Heterogeneous compute management and scheduling
▼
CN Infra
Container Management Cluster and workload management Container Registry Image hosting and integration Device Management Unified hardware device management Cloud Native Network Multi-CNI converged network Cloud Native Storage Container storage and CSI
▼
Ops
Copilot Intelligent O&M assistant
Cockpit Operations data visualization
Billing Center Billing and invoice analysis
▼
OAM
Observability Metrics, logs, and tracing Global Management User permissions and platform settings

As a high-efficiency AI Token production and management system built for intelligent computing centers, Token Factory helps traditional computing centers upgrade to an efficient, profitable Token production and operations model. The platform uniformly manages NVIDIA and domestic heterogeneous computing power, relying on intelligent scheduling and inference optimization to transform distributed GPU resources into low-cost, highly stable, tradable standardized token services. Through the Token Factory manager and unified operations system, it achieves global management of resources, production, costs, and supply-demand, providing capabilities to end users, administrators, operators, and O&M personnel, and driving the intelligent computing center's business positioning to upgrade from "providing compute power" to "producing and operating tokens."

Comments