.png)
.png){kind=icon}
Cloud GPU (vCGPU)
Virtual Machine integrated with one GPU, parts of a GPU or multiple GPUs, specialized for Machine Learning, Video Transcoding and 3D Visualization workloads.
Cloud GPU (vCGPU) provides Virtual Machines (VM) with enhanced computing and graphics processing capability thanks to powerful and versatile NVIDIA GPUs. It is beneficial to organizations and businesses doing their research, development and deployment of Artificial Intelligence (AI) softwares on both testing and production scenarios. It can also speed up Video Rendering, 3D Modelling and Video Transcoding softwares.
vCGPU is built and operated on our multiple Rated 3 – ANSI/TIA942-B Data Centers, also achieving international standard certifications such as ISO 9001: 2015, ISO 27001: 2013, ISO 20001: 2018, ISO 27017:2015.
Connect multiple GPU Virtual Machines
Customer can create multiple GPU Virtual Machines, connect them together in order to divide workload in a Distributed Training model, therefore boosting up processing efficiency.
Loadbalancer & Firewall
Option to buy add-on components such as Loadbalancer or Firewall with reasonable cost to ensure performance as well as system and data protection.
Connect with other Services
Connect Cloud GPU securely with Vcloudia’s other Computing services (Cloud Server, Virtual Private Cloud), Storage services (Cloud Object Storage, File Storage) and Backup service (Cloud Backup) in any way that best serve your needs.
Speed up processing & save time
NVIDIA GPUs help speed up AI training and inference workloads 10-30 times faster compared to only using CPU, therefore reduce processing time from weeks/days to just hours
Quick initialization & flexible scaling
Inherited from Vcloudia’s robust Public Cloud, customer using Cloud GPU can choose to upgrade GPU Virtual Machine anytime needed in a flexible manner, no more wasting time waiting days or weeks as in DIY on-premise model.
Read to use & serve multiple needs
Our GPU Virtual Machine comes with pre-installed OS (Windows/Ubuntu/CentOS), compatible NVIDIA Driver, and CUDA Toolkit. Our service supports all common Machine Learning frameworks and softwares (TensorFlow, Caffe, PyTorch), and can also handle 3D Visualization (3DS Max, Maya) or Video Transcoding (Adobe Media Encoder, Handbrake) workloads.
Deep Learning Training & Inference
Businesses, Organizations and Universities doing their research, development, operation and production of the following types of software/system: FaceID, eKYC, AI Voice Assistant, Chatbot, AI Camera, OCR.
Video Transcoding
Businesses and Organizations providing streaming service (movie, video) for multiple end-user devices; Television station of cities and provinces.
3D Visualization - Rendering
Businesses, Design Studios, Game Studios doing the work of 3D object modelling, video rendering.
Currently, we provide NVIDIA Tesla T4 16GB – one of the most versatile GPU of NVIDIA possessing CUDA cores, Tensor cores and RT cores. It can serve multiple use cases including Deep Learning, Workstation, Video Transcoding. In the near future, we will add to our line of product the newest Ampere GPU cards (NVIDIA A30, NVIDIA A100, NVIDIA Quadro A6000)
Yes, you can in the near future. We are about to launch a whole new vGPU feature, which allows customer to rent only a part of a full GPU cards to optimize usage and cost.
Yes, you can. Customer can create one GPU Virtual Machine with multiple GPU cards attached, or rent multiple GPU Virtual Machines connected to each other in a smooth and secure manner.
Yes, you can. Cloud GPU can connect smoothly and securely with other services such as Cloud Server, Dedicated Private Cloud, File Storage in order to deploy systems like Website, AI, Camera.
Provider
The NVIDIA A30 GPU, powered by NVIDIA's Ampere architecture, Tensor Cores, and Multi-Instance GPU technology, delivers fast, efficient, and secure processing for diverse workloads. This includes large-scale AI inference and high-performance computing (HPC) applications.
The NVIDIA Tesla T4 GPU, built on the Turing architecture, delivers high performance for workloads in cloud computing and inference tasks. Optimized for general-purpose computing environments, the Tesla T4 is equipped with the latest Turing Tensor cores and RT cores for enhanced processing capabilities.
NVIDIA L40S GPU with NVIDIA's Ada Lovelace architecture, next-generation Tensor Cores, and Multi-Workload GPUs deliver fast, secure processing for a wide range of workloads, including large-scale AI interfaces, graphics processing, 3D rendering, and high-performance computing (HPC) applications.