High Performance Computing (HPC) Solutions

Without high-performance computing (HPC) systems, modern research is unimaginable. HPC resources are used to process large amounts of data when fragmented, parallel computations can provide faster results.

The KTU Artificial Intelligence (AI) Centre can provide HPC services.

Here is the architecture of the AI HPC cluster below:

The cluster consists of two computing servers (each: 2x AMD EPYC 7452 32-Core Processor; 2x 256GB (512GB); GPU NVIDIA A100-PCIE-40GB) and home (10TB) and data exchange (10TB) storage using CephFS files. system. The HPC server has Ubuntu 18.04.6 LTS operating system with basic drivers and a SLURM task management system installed.

Each user can upload the required software to their home directory and use SLURM to run the calculations on the HPC nodes (a detailed guide will be provided to users separately).

To use DI HPC services, fill out the form below and we will contact you.

Name, surname(Required)

Contacts (e-mail, phone number)(Required)

Position(Required)

Workplace (department, institute, faculty)(Required)

Description of the purpose of the calculation and the individual software planned to be used (eg. Tensorflow, Mxnet)(Required)

Access time required (maximum 4 months, extended if necessary)(Required)

Planned actual use of resources and storage (approximate, if known)(Required)

CAPTCHA

Name

This field is for validation purposes and should be left unchanged.