Share
We operate two high-performance NVIDIA DGX computing clusters – DGX A100 and DGX H100. Each system integrates eight GPU accelerators with a total GPU memory of 640 GB, providing powerful computing resources for processing and analyzing large-scale datasets.
The two computing nodes are interconnected via InfiniBand with data transfer speeds of up to 400 Gb/s, enabling efficient execution of computationally demanding tasks. In addition to high computing performance, both systems feature a pre-installed and intuitive machine learning software environment, allowing rapid deployment and adaptation to specific projects. The infrastructure is particularly suitable for artificial intelligence, neural network training, machine learning, large-scale data processing, and other computationally intensive applications.
Equipment
NVIDIA DGX A100
- 8× NVIDIA A100 Tensor Core GPUs – 640 GB total GPU memory
- Performance up to 5 petaFLOPS FP16
- 2× AMD EPYC 7742 CPUs – 128 cores in total
- 2 TB system memory
- 28 TB storage
- Price: CZK 100 / GPU / hour
NVIDIA DGX H100
- 8× NVIDIA H100 Tensor Core GPUs – 640 GB total GPU memory
- Performance up to 32 petaFLOPS FP8
- 2× Intel Xeon Platinum 8480C CPUs – 112 cores in total
- 2 TB system memory
- 28 TB storage
- Price: CZK 200 / GPU / hour
