BELLE II: The particle detector to better understand the universe


Theoretical dark matter is one of the most elusive substances in the world, and an important puzzle piece to decode the complexities of the universe’s movement and structure. Even the most practiced Standard Model, which is the golden rule for particle behaviour, does not explain all physics mysteries.

Enter: Belle II is a new particle detector being developed to help researchers try to understand physics beyond the Standard Model, including dark matter. It is an updated and ultra-precise version of the original Japanese Belle detector, which confirmed the existence of other particles that do not fit the Standard Model.

This latest version is likely to collect 50 times more data than the original version (approximately 25 petabytes per year), proving an even bigger challenge than before.

To date, the Belle II experiment is at its third phase and connects around 26 countries and 120 institutions. It deals with the project of a record-breaking collider, named SuperKEKB, investigating some very special particles called B mesons, which spontaneously transform into their own antiparticles and back. Ronald Kotulak, writing for the Chicago Tribune, called the particle “bizarre” and that it “may open the door to a new era of physics” with its proven interactions within the “spooky realm of antimatter”.

The experiment is designed to record data at SuperKEKB and investigate matter-antimatter asymmetries using the B mesons, and as such designed to explore new physics beyond standard practices of particle physics.

B mesons are very special particles, which spontaneously transform into their own antiparticles and back.


The Belle II experiment plans to accumulate more than 100 petabytes (PB) of data, including experimental data and simulations, and process the data over a distributed computing infrastructure across the USA, Italy, Germany, Canada and France. No single supercomputer in the world is able to process that amount of information, due to the high memory requirements that would exceed any resource available in a single location, hence research and education networks and the computing infrastructure they provide become a crucial element to the success of Belle II.

At the moment, the Belle II network infrastructure is based on two main pillars: the SINET 100G Global Ring which connects Japan to the USA and Europe, and LHCONE which accounts for more than 80% of the network’s total computing power and storage resources.

100 PB:

No single supercomputer in the world is able to process that amount of information.


Based on the assumption that while at the time of the Big Bang there were equal amounts of matter and antimatter, today only matter remains. In fact, with perfect symmetry in physics, for every particle created during the Big Bang the relative antiparticle should have been created too. However, a particle and an antiparticle ends annihilating each other with the production of pure energy, so without extremely small violation of the symmetry between matter and antimatter in nature (called “the CP violation”) our universe would not exist as we know it. Belle II can then provide fundamental information about why we do not find antimatter around us, or the imbalance between matter and antimatter, using a B meson.


Transferring vast amounts of data across long distances requires special software tools and really powerful and reliable networks. GÉANT Data Transfer Nodes (DTN) are dedicated high powerful computer servers located in London and Paris, with optimised operating systems and specialised network cards to analyse, test and provide support to long-haul data transfers.

The GÉANT DTN service has therefore become vital for scientists to test data transfer applications, software, protocols, and strategies from their experimental locations to two main central European locations. The GÉANT Research Engagement team liaises with the researchers, helping them to recreate the same software and computing environment in our test locations and then run detailed measurements with them to assess the quality of the data transfer.

R&E networks are then crucial to support experiments like Belle II, where large amount of data have to be shared across a widespread international community.


Belle II has been running systematic Network Data Challenge campaigns, focussed on measuring the maximum achievable bandwidth over the available links across the main data centres of the collaboration. The tests consist of end-to-end data transfers using GÉANT DTN services, which proved key to measure the network speed free from site effects (i.e. grid protocol, speed, LAN configuration, etc). In 2019, a test was conducted to connect the DTN located in London to the Belle II Tier-0 KEKCC in Japan. The Round-Trip time measured between the two sites over the 100Gbps international link was 161 milliseconds!

The test resulted in a performance of a maximum peak of 37Gbps saturating over 92% of the total band with 10 concurrent flows.

“GÉANT DTN services have allowed to verify the connectivity between KEK and EU area minimising the effects induced by sites configuration. Exploitation of DTN capabilities may help in design services for the future data storage infrastructure”

Silvio Pardi, Istituto Nazionale di Fisica Nuclear (INFN)​



2 Exabytes* a year is already carried at high speed and low latency.

* (That’s 2 BILLION Gigabytes!)


Federating services and architecture plus eduroam, the worldwide WiFi roaming service.


Europe-wide procurement frameworks for tailored, market leading cloud offerings.


40+ NRENs, extensive partnerships, programmes, workshops, task forces and events.

The research GÉANT enables touches almost every aspect of our lives.
Not only that, but its networking technology is shaping the internet of tomorrow.

Find out more on the


Skip to content