New supercomputers open avenues to scientific breakthroughs

The computing capacity available to the University of Helsinki is currently being multiplied due to the introduction of several new supercomputers, a development that has the potential to revolutionise a number of disciplines.

Ilpo Vattulainen, professor of biological physics at the University of Helsinki, studies the functional systems formed by biomolecules, or nano-sized molecular machines and intracellular signalling. Such research can be applied to, for example, drug development and, in general, to investigating how a healthy body functions.

Vattulainen’s research is an example of what is known as data- and computing-intensive research, which is currently increasing in prevalence in all fields of science. Data-intensive research means research that requires the utilisation of extremely large datasets whose processing and analysis necessitates increasingly robust computers in terms of computing power and storage capacity.

Data-intensive research often has a link to computing-intensive research, where the computing capacity required for producing results is exceptionally high.

“We simulate cellular activity on the molecular and atomic levels, observing how molecules and the structures they form move about, much like the constant hustle and bustle of a metropolis. This enables us to see, among other things, how a pharmaceutical agent binds with a protein and brings about changes in its functioning, or how certain molecules are able to break up the surface structure of bacteria, thus disabling them. Simulating these processes creates an inordinate amount of data, typically ranging from hundreds of gigabytes to several terabytes,” Vattulainen says.

Ilpo Vattulainen, professor of biological physics at the University of Helsinki, studies the functional systems formed by biomolecules.

Deploying new supercomputers

In addition to its own computing capacity, the University of Helsinki conducts extensive cooperation with both CSC – IT Centre for Science and international service providers.

In 2021 the computing capacity at the disposal of University of Helsinki researchers will grow considerably with the introduction of the LUMI supercomputer maintained by CSC. LUMI will be one of the most powerful supercomputers in the world.

Combined with the recently introduced Mahti and Puhti supercomputers, also maintained by CSC, the University’s computing capacity will grow roughly 20-fold compared to 2019. According to Vattulainen, capacity of this scope is to researchers what a large honeypot is for small bears. His research too will reach a new level.

“Earlier, we were able to investigate something like 10% of the time scale a typical protein requires for doing its job. Soon we will be able to see the entire sequence: how a signal is received and processed by the protein, how the protein modifies its structure as it activates and how it passes the signal on. Now this can be explored down to milliseconds at the atomic level, which has previously been impossible, as computers have not had the capacity to run comprehensive simulations on such processes.”

Researchers must grasp the paradigm shift

Vattulainen says that researchers now need both time for in-depth reflection and training that helps them utilise the increasing computing power.

“Researchers must grasp this paradigm shift. Soon they will be able to focus on entirely new research topics, which can be much more challenging than before, something they wouldn’t even have dared to dream investigating in the past. However, this requires that researchers venture outside their comfort zones and take leaps into the unknown, tackling new and bigger challenges. At the same time, transitioning to a new computing environment requires personal support.”

If this transition is unsuccessful, Vattulainen fears that the new powerful equipment could be underutilised.

“In computational sciences, bigger is always better. For the first time in 25 years, we are taking a genuinely big leap in terms of taking advantage of a new resource.”

Kumpula aims for the top in computational sciences

The IT steering group of Kumpula Campus has published its strategic vision for IT operations as a whole on the campus. The group’s report describes the current state of the campus’s IT operations and the related development needs for the coming years. The vision for development is to turn Kumpula Campus into a global pioneer in computational sciences.

In this, the University is collaborating with, among others, CSC – IT Centre for Science. At the same time, having a competitive and efficient computing capacity of its own is also important to the University.

“It’s important for us in Kumpula to have a solid independent computing infrastructure, enabling us to purchase new equipment and offer new services before they are introduced, for example, by CSC. We can also tailor solutions to local needs. Furthermore, when the infrastructure is in our own hands, we are able to carry out research and provide education in matters related to the infrastructure itself,” says Professor of Computer Science and Vice-Dean Hannu Toivonen, chair of the Kumpula Campus IT steering group.

However, the utilisation of complex computing infrastructures requires knowledge of their software and programming environments. Toivonen points out the environments are also rapidly evolving, which makes it necessary to maintain researchers’ related skills.

“Many researchers can find it too intimidating to start using computing infrastructures and services, but it doesn’t have to be like this. We aim to lower the threshold associated with infrastructures by offering more training to staff.”

The IT for Science team of the IT Centre active in Kumpula maintains the campus infrastructure and provides related services. According to the strategic vision, the team will increase the amount of training offered in the future. Skills will be developed, for example, in a range of training sessions and seminars. Researchers’ networking will be supported through various events to promote the reciprocal exchange of skills.

“What researchers gain from this is an improved ability to utilise different computing services and, thus, conduct research that might not otherwise be possible,” Toivonen says.

Read more:

Researcher, utilise the increasing computing power
  • The resources of the University of Helsinki's high-performance computing are available to all researchers, teaching staff and students. It is advisable for the user to master basic high-performance computing tools such as Linux, Bash, project management systems, and scientific computing software of their field. In the computing environment, work can be performed both interactively and in batches.
  • Equal access to computing resources is managed through FairShare policy. The user can influence the performance of their work by scaling their resource requests as close as possible to the actual need.
  • If a course is organized on campus that requires high-performance computing, it is possible to reserve part of the resources of the computing environment for its use.
  • The university and, for example, the FCCI consortium provide training in the use of computational resources for researchers and teaching staff, as well as for research groups. Upcoming courses are currently announced on mailing lists.
  • I the IT for science team’s HPC Garage consultation sessions the team provides support for the use of computing resources available at the university. You can participate on Tuesdays from 2pm to 3pm via Zoom.
  • More information on HPC Garage
  • Frequently asked questions