GPU as a Service

We operate two highly powerful clusters, NVIDIA DGX A100 and DGX H100. Each is interconnected with eight GPU accelerators and a total memory of 640 GB, providing powerful tools for processing large-scale datasets. Both computing nodes are connected via a network with transfer speeds of up to 400 Gb/s InfiniBand. In addition to high performance, these systems also feature a pre-installed, intuitive software environment for machine learning, enabling easy deployment for your projects.

EQUIPMENT

NVIDIA DGX A100

• 8× TensorCore GPU A100 (total 640 GB RAM)
• Performance: up to 5 petaFLOPS FP16
• 2× AMD EPYC 7742 (128 cores)
• 2 TB RAM
• 28 TB storage array
• Price: 100 CZK / GPU / hour

NVIDIA DGX H100

• 8× TensorCore GPU H100 (total 640 GB RAM)
• Performance: 32 petaFLOPS FP8
• 2× Intel Xeon Platinum 8480C CPU (112 cores)
• 2 TB RAM
• 28 TB storage array
• Price: 200 CZK / GPU / hour