QBO is a container-based cloud computing platform that utilizes Docker containers to replace virtual machines, enabling direct access to CPU, RAM, and GPUs for enhanced performance in AI, machine learning, and low-latency applications. By streamlining the deployment and management of compute instances and Kubernetes clusters, QBO reduces operational costs and eliminates virtualization overhead across various environments.
Funding
Funding not disclosed
Founders
Product
Problem
Traditional cloud computing solutions often rely on virtual machines, which introduce virtualization overhead, increase operational costs, and limit direct access to hardware resources like CPUs, RAM, and GPUs. This can hinder the performance of AI, machine learning, and low-latency applications. Managing compute instances and Kubernetes clusters across diverse environments also adds complexity.
Solution
QBO Cloud is a container-based cloud computing platform that replaces virtual machines with Docker containers, providing direct access to bare-metal resources for enhanced performance. The platform streamlines the deployment and management of compute instances and Kubernetes clusters through container technology, eliminating virtualization overhead. QBO offers a unified AsyncAPI across cloud, on-premises, and air-gapped environments, simplifying infrastructure management. By integrating core infrastructure dependencies such as networking, security, storage, and observability, QBO enables users to build GPU clouds with servers, high-performance network cards, and switches, reducing reliance on external services and lowering costs. The platform supports any application running on Linux or Kubernetes, including Linux and Windows workstations, and is optimized for resource-intensive workloads.
Target Audience
QBO Cloud targets AI labs, platforms, and enterprises that require high-performance computing for AI/ML, real-time processing, and data-intensive workloads, as well as organizations seeking cost-effective and portable cloud solutions.
Features
- Metal-level performance by utilizing Docker containers instead of virtual machines, providing direct access to CPU, RAM, and GPUs.
- Kubernetes-in-Docker (KinD) and Docker-in-Docker (DinD) support for isolated and reproducible environments.
- Integrated infrastructure dependencies, including DNS, Load Balancing, Firewall, Authentication, and ACME.
- AsyncAPI for dynamic compute provisioning, real-time API communication, and seamless scaling across distributed environments.
- CNCF conformance, ensuring adherence to cloud-native computing best practices and interoperability with other conformant Kubernetes systems.
- Hardware and architecture-agnostic, running on any x86 or ARM server without specialized hardware or virtualization layers.
- Support for NVIDIA GPU Operator and NVIDIA Container Toolkit for GPU-accelerated workloads.