Mahti - Services for Research
Mahti is a supercomputer designed for massively parallel jobs that require large floating point performance and a capable interconnect. Mahti is in particular geared towards medium to large scale simulations. Also smaller parallel workloads that are able to use full nodes efficiently can utilize Mahti.
- Total 1404 CPU nodes equipped with powerful AMD Rome CPUs, all connected with a fast interconnect. This allows jobs to scale across the full system. Each node has 128 cores running at 2.6 GHz, with 256 GiB of memory
- Total 24 GPU nodes equipped with AMD Rome CPUs and four Nvidia Ampere A100 GPUs, 512 GiB of memory and fast local Nvme disks (3.8 TB)
- The possibility to reserve full nodes or GPUs so that your jobs can extract full performance from each node
Using the HPC environments requires elementary Linux skills. We recommend you are familiar with basics of Linux command line usage before starting.