> For the complete documentation index, see [llms.txt](https://quantumgrid.gitbook.io/quantum-grid/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://quantumgrid.gitbook.io/quantum-grid/hardware/virtual-machines-vms.md).

# Virtual Machines (VMs)

Virtual Machines (VMs) are the backbone of our service, allowing users to rent and run different operating systems or applications on the same physical hardware. We support several types of VMs to cater to varied computational needs:

Our Virtual Private Server (VPS) services, powered by the **Microsoft Hyper-V virtualization platform**, offer a variety of VM types to meet diverse computational needs. With Hyper-V's robust performance and reliability, users can experience smooth, efficient operations for both **Linux** and **Windows** servers. Hyper-V ensures better control and isolation compared to shared hosting solutions, making it an ideal choice for users seeking reliable and scalable VPS solutions.

We support several types of VMs designed for specific tasks:

* **Type-1 Hypervisor VMs (Bare-metal VMs)**

These VMs run directly on the hardware, ensuring near-native performance. They are perfect for **compute-intensive applications**, such as deep learning, high-frequency trading, and resource-heavy enterprise applications. Leveraging **Hyper-V**, these VMs offer greater isolation and control compared to shared hosting solutions, providing dedicated resources to each virtual machine.

***

* **Type-2 Hypervisor VMs (Hosted VMs)**

Running on top of an existing operating system, these VMs provide flexibility and ease of management while sacrificing a small portion of performance. They are ideal for **general-purpose tasks**, development environments, and testing. The flexibility offered by Type-2 Hypervisor VMs makes them an excellent choice for businesses with evolving computing needs. Hyper-V ensures efficient resource allocation, making this a solid choice for many users.

***

* **Containerized VMs**

Using technologies like **Docker** and **Kubernetes**, our containerized VMs are lightweight and portable, delivering exceptional **performance and resource efficiency**. These containers are perfect for **microservices architectures** and **cloud-native applications**, where rapid scaling and minimal overhead are crucial. Containerized VMs allow for efficient orchestration of multiple application components, ensuring seamless scaling and deployment.

***

* **GPU-accelerated VMs**

For tasks requiring advanced graphical power, such as **AI/ML training**, **scientific simulations**, and **rendering**, we offer **GPU-accelerated VMs**. Powered by **NVIDIA vGPU** and **CUDA**, our platform enables direct access to the host's GPUs using **GPU passthrough**, ensuring unmatched performance for specialized tasks. This feature allows virtual machines to directly access the host's GPU, offering significant power for **machine learning**, **data analysis**, and **graphical tasks**. We are currently expanding our GPU-accelerated offerings to include even more powerful capabilities, supporting the growing demand for **high-performance computing** and **graphic-intensive applications**.