NAISS
Objectives
Let’s step up and have an overview of NAISS
NAISS HPC Clusters
https://www.snic.se/resources/compute-resources/
Mostly Intel-based CPU:s and Nvidia GPU:s
UPPMAX:
Rackham - General computational resource and also focused towards Life Science
Bianca - NAISS-SENS, resource dedicated to research on sensitive data
PDC (KTH):
Dardel - General computational resource
Not intel/Nvidia but Cray/AMD CPU:s and GPU:s
NSC, Linköping:
Tetralith - General computational resource
C3SE, Chalmers:
Alvis - Accelerator-based resource dedicated to research using AI techniques AND the international
LUMI, Finland
LUMI - General computational resource, aimed at larger projects in Sweden
Not intel/Nvidia but Cray/AMD CPU:s and GPU:s
Locally financed systems for local users
UPPMAX
Snowy - GPU:s and long runs
LUNARC, Lund
Aurora
C3SE, Chalmers
Vera
NSC, Linköping:
BerzeLiUs - for AI and machine learning
HPC2N, Umeå:
Kebnekaise - General computational resource
Swedish Science Cloud (SSC)
Complement, offering “advanced functionality” to users who:
need more flexible access to resources
e.g. more control over the OS:s and software environments
want to develop software as a service
want to explore recent technology such as for
IoT applications
“Big Data” e.g. Apache Hadoop/Spark
The UPPMAX cloud Dis (Swedish word for “haze”) was introduced in October 2017 and upgraded during 2020.
East region in the Swedish Science Cloud https://www.uppmax.uu.se/resources/systems/the-uppmax-cloud/
Other clouds
Umeå University (North, HPC2N).
Chalmers University (West, C3SE).
STORAGE
SNIC Storage: Swestore
The purpose of Swestore allocations, granted by Swedish National Allocations Committee (SNAC), is to provide large scale data storage for “live” or “working” research data, also known as active research data.
The projects
No cost to researchers all over Sweden.
3 different levels regarding storage and core hours per month
https://uppmax.uu.se/support/getting-started/applying-for-projects/
Compute at UPPMAX
Small: 2000 cpu-hrs/month and 128 GB of storage.
Can be increased to 10,000 hrs
PI: PhD student or higher.
Medium: 10,000-200,000 cpu-hs/month
PI: Assistant professor (forskarassistent) or higher.
Large 100,000 cpu-hs/month
PI: Assistant professor (forskarassistent) or higher.
Storage at UPPMAX
Small: 128 GB-10 TB of storage.
Can be increased to 10,000 hrs
PI: PhD student or higher.
Medium: 10-100 TB
PI: Assistant professor (forskarassistent) or higher.
Large: More than 100 TB
PI: Assistant professor (forskarassistent) or higher.
Application rounds
https://www.naiss.se//#application-rounds-for-compute-and-storage-resources