Computational research is one of the focus areas in Aalto University, and Aalto Scientific Computing makes that possible.
The Science-IT project was founded in 2009 (with roots going back much further) and has since expanded from high-performance computing services to a complete package: we provide computation, data management, software, and training. Our partnerships with departments and central IT services allow a streamlined experience from personal devices to the largest clusters.
To reflect our expanded services, we have rebranded to Aalto Scientific Computing to reflect our greater mission and partners.
Many Centres of Excellence and departments at Aalto University are using our resources with great success. There are currently over 1000 user accounts from all six different schools and at least 14 different departments using our resources. Science-IT is administered from the School of Science with additional university-level funding - our HPC services are available to all Aalto University, free of charge.
.. toctree:: :maxdepth: 1 for-proposals join
We don't just provide computing hardware, but a complete package of infrastructure, training, and hands-on support. All of these three activities feed back into each other to improve the whole ecosystem.
We provide many types of services:
Aalto Scientific Computing serves as a hub of computational science at Aalto. We guide researchers to the right service, regardless of who is providing it.
Science-IT serves as the coordinator, and runs the Triton cluster, the physical hub of large scale computational and data-intensive research at Aalto. As such, we maintain many active collaborations which allow us to guide researchers to the right resource, regardless of who provides it.
.. toctree:: :maxdepth: 2 science-it
Science-IT is the organizational manifestation of Aalto Scientific Computing.
Science-IT concentrates on mid-range computing and special resources needed by researchers in the School of Science. With local resources, we can provide high-quality support and even research-project-level customization. Because our resources are integrated into the Aalto IT environment, with regular local training in the scientific computing practice to entry-level users, our resources enjoy an ease of access and lower barrier to entry than, for example, CSC HPC resources. We are also a basic research infrastructure, enabling the integration of separately purchased resources to our cluster and storage environments, with dedicated access for the purchaser.
Our team is mainly known for providing the :doc:`Triton cluster </triton/index>`, a mid-range HPC cluster with ~10000 CPUs, 5PB storage capacity, Infiniband network, and ~150 NVIDIA GPUs for deep learning and artificial intelligence research. We provide a :doc:`Jupyter Notebook </triton/apps/jupyter>` based interface to enable light computing with less initial knowledge required to make our services easily accessible to everyone. Our team also works with the CS, NBE, and PHYS departments to provide :doc:`HTCondor </aalto/htcondor>` (high throughput computing), data storage, and a seamless computational research experience. We maintain http://scicomp.aalto.fi, the central hub for scientific computing instructions and have a continuous :doc:`training program </training/index>`, :doc:`Scientific Computing in Practice </news/index>`.
These departments are members of Science-IT, and their local IT staff provide a great deal of scientific computing support, and in fact all the Science-IT team above is contained here. These departments resources are seamlessly integrated with Aalto's HPC resources.
.. toctree:: :maxdepth: 1 cs-it
We are a leading member of the Finnish Grid and Cloud Infrastructure (FGCI), a university consortium to support mid-range computing in universities. FGCI, via Academy of Finland research infrastructure grants, funds a large portion of our work. Thus, we maintain ties to most other universities in Finland as well as CSC, the national academic computing center. Through the FGCI, we provide grid computing access across all of Finland and Europe.
Our team overlaps with the Departments of Computer Science, Neuroscience and Biomedical Engineering, and Applied Physics. The IT groups in these departments provide advanced Triton support.
We maintain close collaboration with Aalto University IT Services (ITS). We are not a part of ITS, but work closely with them as the computational arm of IT Services. ITS provides the base which we repackage and build on for many of our services.
Our team maintains ties to Aalto Research and Innovation Services to guide data and research policy. Triton is an Aalto-level research infrastructure. Our staff is involved in research policy making, including ethical, data security, and data management. Our team contains several Aalto Data Agents.
We partner with CodeRefinery, a Nordic consortium to assist in training of scientists, to provide training and support computational competence.
This table lists people supporting Scientific Computing at Aalto University who considers themselves a part of ASC. If you want to be added here, let us know. We welcome all contributors. There is no Aalto Scientific Computing, just people who want to make computing better.
| Affiliations | Specialties | |
| Richard Darst | Science-IT, CS-IT, Data Agents, Aalto RSE | Data science, Triton, teaching, usability |
| Ivan Degtyarenko | Science-IT, PHYS-IT | Triton, HPC hardware, HPC OS, teaching |
| Enrico Glerean | Science-IT, NBE, Data Agents, Aalto ethics committee | Triton, ethics and personal data, data. |
| Simppa Äkäslompolo | Science-IT, PHYS-IT | Triton, HPC OS, parallel software |
| Jarno Rantaharju | Science-IT, Aalto RSE | Software Development, HPC software and optimization, profiling |
| Marijn van Vliet | Science-IT, Aalto RSE | Software development, data, neuroimaging |
| Essi Heikkinen | Science-IT | Natural language processing, teaching, documentation |
| Simo Tuomisto | Science-IT, CS-IT | Software Development, HPC software design and optimization, GPU computing |
| Mikko Hakala | Science-IT, CS-IT, NBE-IT | Triton, data storage systems, HPC administration |
FCCI Tech is a seminar serious about how scientific computing/research computing groups run. We present the things we have learned and how we run internally. This is targeted towards other academic technical specialists, but anyone is welcome to attend and learn.
.. toctree:: :maxdepth: 1 /tech/index
Most of the computationally-intensive research outputs from our member departments use our resources. In addition, at least the CS and NBE departments use our data storage for most big data projects. You may view the our research results using research.aalto.fi (Science-IT infrastructure section).
Our users come from countless research areas:
- Method development
- Computational materials research
- Network research
- Neuroscience
- Data mining
- Deep learning and artificial intelligence
- Big data analysis
.. toctree:: web-accessibility ../README