What is GPUBox?
GPUBox is a virtualization technology and a cloud-ready GPU-computing platform that provides a completely new angle on the usage of the GPUs. It is the foundation of a fully-functioning, scalable and elastic cloud infrastructure.
The GPUBox software simplifies GPU management by separating the application and operating systems from the underlying GPU devices. It is a solution that allows the dynamic sharing of GPU devices from the same pool, by many users. The GPU-cloud-enabled infrastructure can be set up, regardless of its size and purpose, by small 3D graphics studios as well as larger providers. Also, the GPUBox infrastructure can be extremely efficient thanks to recent network technology with its high throughput and low latencies.
GPUBox enables on-demand provisioning of GPU devices to a physical or virtual machine with a Linux or Windows operating system. The pool of GPU devices is shared among users which leads to reduction in the total power consumption and idle-running hardware.
Basic usage of the GPUBox software is easy to understand.
For example, to allocate 20 GPU from the pool of devices to the Client, the user
issues a simple command in the terminal:
$ gpubox add 20
After that, the supported applications will recognize 20 GPUs as if they were installed in the system. When the GPUs are no longer needed, the user can easily remove them from the system by using the command:
$ gpubox drop all
With GPUBox, the GPU cloud infrastructure is not limited only to the most expensive professional devices. GTX-class cards can be easily applied.
The first implementation of the GPUBox technology is GPUBox. It responds to the needs of today's 3D graphics purposes by leveraging CUDA-based renderers with GPU cloud computing.
The next implementation of GPUBox will be GPUBox Scientist, crafted especially for
applications that might be used for such computations as:
- molecular and fluid dynamics
- computational chemistry
- numerical analytics
- structural mechanics
- computational finance
- weather and climate simulations
- materials science
- seismic simulations
Sharing the same device among many users
Any GPU working within the GPUBox infrastructure can be allocated to the system in the shared usage mode, which means that the GPU can be used simultaneously by many users.
Dynamic provision of any number of GPU devices
GPUBox enables users to allocate the GPUs on demand. In less than a second, any number of the GPU devices can be allocated to the system ready to be used.
Intuitive Web Console
Convenient browser-based user interface that can be used for administrating, monitoring and allocating/dropping the GPUs.
Scalability and flexibility
New GPU devices can be seamlessly introduced to the GPUBox infrastructure and provided to the users. Also, according to the current needs, the infrastructure can be easily reconfigured.
No modification is necessary in the existing
Supported applications see the allocated GPUs as if they were physically mounted in the PCIe slots.
Reduction of total cost and better GPU utilization
With the GPUBox technology a smaller number of GPU devices can serve a larger number of users. Eventually it also leads to lower power consumption.