DaoCloud

DaoCloud, an innovative leader in cloud-native AI, leverages its continuous open-source contributions and deep technical expertise in cloud-native technologies to launch d.run - a computing power scheduling platform. d.run efficiently manages heterogeneous computing resources, seamlessly integrates mainstream LLMs, and enables flexible scheduling and sharing of compute resources, accelerating AI application innovation across industries.

d.run provides:

1. Model Engineering Platform
   An end-to-end toolchain covering model development, training, inference, deployment, and application. It accelerates AI delivery from concept to production, supports diverse scenarios from enterprise AI assistants to LLM services, and offers enterprise-grade capabilities including on-premises deployment, multi-tenancy, access control, and metering.
2. Intelligent Computing Orchestration
   Optimises compute, storage, networking, and inference performance for high-concurrency AI workloads. With intelligent scheduling, GPU virtualisation, and software-defined optimization, it maximises GPU utilisation (80%+), improves operational efficiency, reduces latency and cost, significantly improves training efficiency and LLM inference throughput (up to 24×), and enables sustainable AI.
3. Multi-cloud Heterogeneous GPU Management
   Provides unified pooling and scheduling of heterogeneous GPUs across multi-cloud and multi-cluster environments. It supports GPUs from multiple vendors, enables fine-grained dynamic allocation for training and inference, avoids vendor lock-in, protects ROI, optimises TCO, and ensures compliance.

Since its establishment, DaoCloud has remained committed to driving enterprise digital and intelligent transformation, serving industries including financial services, automotive, advanced manufacturing, energy, retail, education, and government.

DaoCloud has several CNCF Ambassadors, ranks among the global leaders in Kubernetes project contributions, and includes members of the Kubernetes Steering Committee and the Kubernetes Security Response Committee. It has initiated and contributed to multiple open-source projects in the CNCF Sandbox and serves as a maintainer of several widely adopted projects.