Our Features
  • Top-grade Tesla GPUs: V100 32GB, V100 16GB, P100 16GB
  • High security due to Bare metal access. Only one customer reside on physical server with root access
  • High interconnectivity: 56GB/s Ethernet/Infinityband and NVLINK
  • Full performance due to no virtualization overheads
  • Per-second billing
  • We accept crypto: Bitcoin, Ethereum payments
Get free trial to test performance
  1. Register new account
  2. Fill in a form to request trial access
Rent a GPU server for DeepLearning
  1. Register new account
  2. TopUp your balance
  3. Choose GPU server and start rental
Instantly Available Bare-metal GPU Servers
4x V100 32GB SXM (NVLINK) x16x16x16x16

3.72 USD/h

2xE5-2698RV4, Cores: 40/80, RAM: 256GB

Storage: NVME, 1600Gb

Internet: 1Gbit shared

Id:9

Rent
4x RTX 3090 x16x16x16x16

3.1 USD/h

2xE5-2698RV4, Cores: 40/80, RAM: 128GB

Storage: SSD, 400Gb

Internet: 1Gbit shared

Id:13

Rent
Available by request GPU Servers and Cluster
20-100x P100 Cluster

13 USD/h

4x P100 16GB SXM (NVLINK)

E5-2698RV4, 64GB RAM, SSD400GB,

LAN Ethernet or Infinityband 40Gbit, Internet: 1Gbit shared or dedicated

Id:1

Request Cluster

For long term or high resource volume rentals we provide manually preconfigured servers and clusters with specified image and hardware configuration. Our infrastructure includes onehost Nodes with NVLINK (Up to 4 GPU in NVLINK), RDMA over Infinityband (up to 100gbps) with up to 30 Nodes in network, NAS with up to 100Tb space. We provide help in configuration of Hadoop Yarn GPU cluster for Bigdata processing or DeepSpeed distributed Deep Neural Network training

Request custom GPU Server or Cluster