SURF.net news

Category: Uncategorized

  • Successful trial: 1.2 Tbit/s data transfer to prepare long-haul network for AI and supercomputing

    Successful trial: 1.2 Tbit/s data transfer to prepare long-haul network for AI and supercomputing

    CSC, SURF, and Nokia have successfully tested a high-capacity, quantum-safe fibre-optic connection exceeding 1.2 terabit per second (Tbit/s) between Amsterdam and Kajaani (Finland) with data traversing over 3.500 kilometres. The trial demonstrated the potential of ultra-fast, cross-border connectivity for research.

    Tests were carried out along several routes, including the longest, which spanned 4,700 km through Norway at a capacity of 1Tbit/s. To put this in perspective, 1 Tbit/s is enough to stream 200,000 full HD movies (at 5 Mbit/s each) simultaneously.  

    These results are particularly promising as the research community prepares for supercomputers and AI Factories to come online – where reliable, scalable, and secure connections will be critical to supporting some of the world’s largest datasets and most demanding workloads.

    The test used a combination of real research data and synthetic data, transferred directly from disk to disk – from SURF’s facility in Amsterdam to CSC’s data center in Kajaani, across five production research and education networks: SURF (the Netherlands), NORDUnet (Nordic backbone), Sunet (Sweden), SIKT (Norway) and Funet (CSC’s network in Finland). 

    The network solution was based on Nokia’s IP/MPLS routing and quantum-safe optical networking gear. Nokia’s IP technology successfully demonstrated Flexible Ethernet (FlexE) to accommodate “elephant flows”, or very large continuous flows of data, and its high-capacity optical transport technology showed the ability to handle massive data sets generated by HPCs over long distances. 

    Distance between supercomputing centers is no obstacle

    With the exponential growth of research data, especially for training large-scale AI models, the need for resilient, high-throughput and secure connectivity is more critical than ever. This test confirms that multi-domain, high-capacity data transfers across European research networks are both feasible and future-ready. Testing an operational network connection over long distances provides unique insights into data transport and storage of large data volumes. The tests are crucial for improving the infrastructure for data-intensive research.  

    “We design research networks with future needs in mind,” said Jani Myyry, Senior Network Specialist at CSC. “CSC’s datacenter in Kajaani already hosts the pan-European LUMI supercomputer, and with the upcoming LUMI-AI supercomputer and AI Factory coming online, reliable and scalable data connections throughout Europe are essential. Even though the geographical distance is significant, it poses no obstacle to data traffic.” 

    Ready to take the next step in aligning European supercomputers

    “As SURF, we are ready to take the next step in aligning European supercomputers,” said Arno Bakker, Senior Network Specialist at SURF. “These efforts offer future perspectives to train federated LLM as OpenEuroLLM

    on LUMI and Snellius or for a researcher to compute on LUMI with very large datasets hosted at SURF, such as the KNMI (The Royal Netherlands Meteorological Institute) datasets.” 

    “This test would not have been possible without the support of NIKHEF, the Dutch national institute for subatomic physics. Their experience in handling huge amounts of physics measurement data and available hardware ensures that these kinds of tests can be done successfully. In addition, Bakker says we are very grateful to our Scandinavian partners for their help setting up this trail connection. This is again an example of the continued good cooperation between NRENs to create the best possible international infrastructure for research and education.”

    “Groundbreaking trials like this highlight how advanced networks are foundational to unlocking the full potential of AI and high-performance computing,” said Mikhail Lenko,Customer Solutions Architect for Nokia. “This successful collaboration with CSC and SURF is a testament to the innovation and leadership of the scientific community, and to what’s possible when we work together.”

    “As the network prepares for the next wave of supercomputers and AI Factories, we are proud to deliver the quantum-safe, high-capacity, and resilient IP/MPLS and optical infrastructure that makes these systems viable. We look forward to continuing our support for global research and education networks, helping them scale with confidence and drive the next generation of discovery and innovation.”

  • Towards a European Time & Frequency Network

    Towards a European Time & Frequency Network

    Introduction

    Last month, GÉANT organized a physical edition of the Special Interest Group Time & Frequency Network (SIG-TFN) at the Joint Science Centre (a research and advisory body of the European Commission). During this event in Italy, NRENs and NMIs came together to further develop the framework for a European Time & Frequency network. In this blog, I will explain what this network aims to achieve and what role SURF plays in it.

    From National to International

    In recent years, many national NRENs (National Research & Education Networks), together with NMIs (National Metrology Institutes), have established Time & Frequency networks. These networks distribute time and frequency signals via optic networks to researchers and, in some cases, commercial parties. Various types of scientific research benefit from the improved synchronization and calibration of measuring equipment, leading to more refined results.

    Want to learn more about how we do this and how it works? Read more about SURF Time & Frequency [here], or listen to the podcast I recorded on this topic with SURF.

    What was still missing was cross-border connectivity between these national networks. GÉANT is now working with NRENs and NMIs to develop a network that connects individual countries.

    A Golden Partnership: NRENs & NMIs

    In recent years, NRENs and NMIs have increasingly found common ground and are working together more closely. NMIs provide the clocks and the sources for the Time & Frequency signals, while NRENs supply the network infrastructure and expertise to distribute these signals.

    Interestingly, the creators of these atomic clocks—who typically provide the source for current Time & Frequency networks—now require the network themselves to further develop the so-called optical clocks.

    Optical Clocks

    Optical clocks are the next generation timekeeping devices, capable of measuring time with far greater precision than the current cesium atomic clocks. By using lasers instead of microwaves (higher frequency means higher precision), these clocks can measure time up to hundreds of times more accurately than the current generation. By 2030, this technology is expected to lead to a redefinition of the second.

    But how do we know if an optical clock is working correctly? Measurement requires comparison. If you want to measure the length of a bacteria, you wouldn’t use a standard tape measure—you need a more precise instrument. Similarly, an optical clock, which would only drift by one second over 15 billion years, can only be tested by comparing it to another optical clock.

    However, you can’t simply pack an optical clock into a suitcase and take a train from Amsterdam to Braunschweig (where another optical clock is being developed). A different approach is needed.

    A picture of a research clock at the UvA. More info see iqclock.eu.

    Comparing Clock Signals via the Network

    What is possible, however, is the transmission of an optical clock’s frequency signal via a network. This allows clocks in different locations to be compared. And not just two clocks—you need multiple clocks to determine if any of them have a deviation.

    The technology for transmitting these frequency signals is now so advanced that the signal loss is smaller than the uncertainty of the clocks themselves. This is precisely what the first phase of the Core Time/Frequency Network (C-TFN) enables. Through this network, Dutch clocks at UvA Amsterdam Science Park are compared with clocks from Germany and France. Amsterdam Science Park will play a central role in this new network, creating a unique situation where signals from the world’s best NMIs converge and are analysed.

    National Ultra Stable Optical Frequency Network

    Not only timekeeping researchers benefit from these signals. Several SURF members, such as ESA, VU, and TU/e, have already expressed interest in gaining access to these highly precise frequency signals.

    The technology enabling this, Ultra Stable Optical Frequency Transfer (a kind of “White Rabbit on steroids”), is another factor of 1000 more precise than White Rabbit, reaching precision levels in the pico- and femtosecond range. This has applications in fields such as quantum computing. That’s why we are already installing filters in the national network to distribute these signals. This way, SURF continues to lead the way in supporting researchers in the Netherlands.

  • New technology for lightning-fast data traffic: SURF and ASTRON realise 400G connection with OpenZR+

    New technology for lightning-fast data traffic: SURF and ASTRON realise 400G connection with OpenZR+

    For more information see: https://www.surf.nl/en/news/new-technology-for-lightning-fast-data-traffic-surf-and-astron-realise-400g-connection .

    Scientific research in the Netherlands has gained an advanced networking technology: OpenZR+. Together with ASTRON, the Dutch institute for radio astronomy, we used this technology to create a 400G connection that can transmit even sharper images of space. Institutions in need of scaling up network connectivity can easily request this within SURF’s All-In network tariff.

    Since July 2024, SURF members who need a network upgrade for their research can request additional bandwidth or additional services from us within the All-In network tariff. For ASTRON, we thus realised a direct connection between the research facilities in Groningen and Dwingeloo, with increased bandwidth from 10 to 400 Gigabits per second.

    Faster data transfer LOFAR telescope

    For astronomers working with the LOFAR telescope, the network upgrade means they can now study the universe in even more detail. LOFAR (Low Frequency Array) is the world’s largest radio telescope operating at the lowest frequencies observable from Earth. The telescope consists of thousands of small antennas scattered across Europe.

    Thanks to the improved network connection, LOFAR can now send more signals simultaneously to a central computer (the ‘correlator’), which processes the data and turns them into images of space. Thanks to the increased bandwidth, this is done not only faster, but also in higher quality, resulting in even sharper LOFAR images.

    Collaboration strengthens research

    In addition, the new network connection allows astronomers to sign up for multicast streams of data coming directly from the LOFAR stations. “This offers researchers direct, real-time access to the latest scientific data, at the moment they need it,” says Julian Kootstra, network engineer at ASTRON. “Having your own cluster gives you instant access to new data – a revolutionary way to push the boundaries of our knowledge.”

    Paul Klop, optical network architect at SURF, is also pleased with the result: “With this 400G connection, we are supporting researchers at ASTRON to achieve their scientific goals. The cooperation with ASTRON was excellent and shows how we can work with our members to deploy innovative technologies to enable groundbreaking research.”

    Advantages OpenZR+ technology

    The network upgrade at ASTRON used 400G-ZR optics, a technology that amplifies the signal in the network without separate transponders. This requires fewer intermediate components, reducing the complexity of the network. This provides several advantages:

    • Less chance of malfunctions
    • Simpler network management
    • Lower costs through more efficient equipment
    • Reduced power consumption, contributing to sustainable data processing
    • More compact equipment, allowing data centres to be set up more efficiently

    In SURF’s network of the future, SURFnet-Infinity, OpenZR+ technology will be the standard for all medium-distance connections.

    SURFnet-Infinity: the future of research networking

    The 400G upgrade is part of our broader strategy to set up the network of the future: SURFnet-Infinity. This network, based on open standards, allows equipment from different suppliers to be connected over long distances. This makes the SURF network more flexible and future-proof.

    Through the All-In network tariff, we aim to provide educational and research institutions with network solutions tailored to their specific needs. In this way, we can support the Dutch research community even better.

  • Workflow Orchestrator Programme – Codesprint

    Workflow Orchestrator Programme – Codesprint

    Last week SURFs network automation team spent a fruitful week in Berkeley (CA) together with colleagues from ESnet and Géant. As workflow orchestrator partner members, SURF, ESnet and Géant are committed to the workflow orchestrator software ecosystem and collaborate on creating a useful set of tools for parties who are interested in automating their network.

    First codesprint

    As users of the orchestrator software we share the need to continuously improve and add more features. The coming year SURF, ESnet and Geant plan to execute 12 sprints to work on the backlog of issues. This will be done by forming a (virtual) team of software engineers who will together work on issues that have been identified as important. The codesprint represents an important step in sharing knowledge and kickstarting this process.

    Results

    The codesprint results in a nutshell:

    • 22 issues closed with a combined issue weight of 45
    • 7 issues in progress with a total weight of 23
    • 2 releases including one Major release candidate. version 2.10, version 3.0.0rc1
    • Numerous new contributors
    • Many good stories and lots of laughs!


  • Field trip to visit our colleagues at DFN-Verein

    Field trip to visit our colleagues at DFN-Verein

    What an insightful day in Berlin! A big thank you to Stefan Piger and Leonie Schäfer from DFN-Verein for the engaging discussions and valuable learnings. It was a pleasure to exchange ideas and explore opportunities for collaboration together. Looking forward to continuing this conversation!

  • Digital lead Netherlands under pressure

    Digital lead Netherlands under pressure

    Traditionally, the Netherlands has been an important digital hub in Europe. Sea cables landed in the Netherlands and made Amsterdam a digital hub. That had enormous appeal for parties that need a strong digital network, such as education and research. The leading position that the Netherlands has had for a long time is now under pressure. Why is that? Why is it bad if we lose that position? And what will it take not to lose this important position?

    Listen to this episode (in Dutch) of SURFsounds with Peter van Burgel, CEO of Amsterdam Internet Exchange (AMS-IX) and Alexander van der Hil, international policy and strategy advisor at SURF.

    https://podcast.surf.nl/@SURFsounds/episodes/digital-copposition-dutch-under-pressure-nuzey

    https://open.spotify.com/show/6IcYxQzB6wCCvxFJL34gzM

    https://podcasts.apple.com/nl/podcast/surf-sounds/id1682253126

    https://soundcloud.com/surf_sounds/digitale-koppositie-nederland-onder-druk/s-SdMgOCkefJM?si=140fbcd731d549088f76a46ff4fd0d87&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing

    Via: https://www.surf.nl/podcast/digitale-koppositie-nederland-onder-druk

  • Network dashboard goes electric!

    Network dashboard goes electric!

    Last summer, a huge transition was implemented under the hood of the Network Dashboard. At first glance, little seems to have changed, but especially if you click around, you will now notice that it works much faster. The major overhaul of the network dashboard had been on our wish list for a long time, because the complex authorization rules made the Network Dashboard feel like an old diesel at times (or even slower!).

    The architecture of the Network Dashboard has been completely overhauled by no longer collecting all data from different systems (orchestrator/ipam/CMDB/CRM/jira) in real time, but by preparing all static data in advance in a document in the replica set. As a result, we only have to make a few very fast calls instead of 500 separate API calls. Only the traffic graphs and health status of the services are now retrieved live from the influx database.

    How quickly is such a replica set updated?

    The preparation of these documents in the replica set is triggered with every change to a subscription that the Workflow Orchestrator (see workfloworchestrator.org) executes on a service. The replica is also refreshed every night. This allows us to guarantee the data integrity between the orchestrator and the replica set. Only in the case of real-time changes on the network or due to self-service actions will the replica set briefly lag behind the actual state as recorded in the Workflow Orchestrator. For example, adjusting a customer alias is not immediately visible in the network dashboard, but must be processed in the replica set. Simple changes take ~4 seconds, while larger changes can take many times longer on average. At the moment we are still working on improving the automatic refresh of the pages after self-service actions, until then you will have to manually refresh a page after a change to see the latest up-to-date information.

    SURFdomeinen will be available in the network dashboard early next year. Due to the complex migration of domain names and zones, this will be carried out in phases. The product manager of SURFdomeinen will communicate about this in due time.

    Finally, we have introduced a new look-and-feel with a renewed landing page, on which all network services of the standard network portfolio are clearly presented in one tile.

  • Stronger digital ties across the Atlantic

    Stronger digital ties across the Atlantic

    Stronger digital ties across the Atlantic


    Dutch, Nordic, and Canadian research and education networks
    upgrade their trans-Atlantic link to 400 Gigabits per second (Gbps) as
    part of the Advanced North Atlantic (ANA) consortium.

    The research and education network organizations of the Netherlands, the Nordic countries
    and Canada have signed an agreement to upgrade the existing connection between
    Amsterdam and Montréal from 100 Gbps to 400 Gbps, placing it among the most powerful
    intercontinental connections in the world.

    The Amsterdam-Montréal connection was established five years ago by SURF, the national research and education network (NREN) of the Netherlands, in collaboration with NORDUnet, the regional NREN of the five Nordic countries. With the new agreement, CANARIE – the federal partner in Canada’s NREN – joins, with the three partners contributing equally.

    The Advanced North Atlantic (ANA) consortium is a joint effort of nine research and education networking organizations in North America, Europe, and Asia committed to maintaining, operating, and sharing a high-speed trans-Atlantic network infrastructure.

    Partners chip in

    The strength of ANA consortium lies in the contribution that each partner makes to the system, and shares this with the ANA membership to achieve a better infrastructure for all. Since all members are organizations that exist to build and operate infrastructure to accommodate the needs specific to research and education, the willingness to contribute has been high.

    “This upgrade is a natural response to the increasing demand of research and education data and services exchange across the Atlantic Ocean, for example research enabled by the Large Hadron Collider (LHC) and Square Kilometer Array (SKA) instruments. The ANA consortium is an excellent example of the impact of partnerships on the collective good. This partnership enables our three national research and education networks to do more together than we can do on our own, maximizing the impact of our individual investments – both on the researchers in our home countries and the global science initiatives they collaborate in,” says Mark Wolff, Chief Technology Officer, CANARIE.

    “The partners chip in when and where they can, but always with an eye to the capacity and redundancy of the entire system. This is one of the advantages of being in the research and education network community. Collaboration is smooth and trustful,” says Lars Lange Bjørn, Team Lead, Network & Service Technology, NORDUnet.

    A connection landing in Canada

    ANA began a little more than a decade ago as a pilot project, trying to probe interest in a 100 Gbps connection for research and education across the North Atlantic Ocean. In 2013, the first such connection was established.

    Currently, 13 connections across the Atlantic Ocean are operated as a unified system by the ANA partners.

    “Notably, the other 12 connections all land in the USA. Our connection is the only one landing in Canada, and this is essential for sufficient fail-over capacity and redundancy in the system,” says Harold Teunissen, Director Network and Campus at SURF.

    The 400 Gbps upgrade for the Amsterdam-Montréal connection is expected to be completed by September 2025.

    The Advanced North Atlantic (ANA) consortium members are KISTI, SURF, NORDUnet, NII,
    Indiana University, ESnet, Internet2, CANARIE, GÉANT.

    About NORDUnet

    NORDUnet is a collaboration between the National Research and Education Networks (NRENs) of the five Nordic countries, i.e., Denmark (DeiC), Finland (Funet/CSC), Iceland (RHnet), Norway (Sikt), and Sweden (Sunet). NORDUnet operates a world-class data network, based on dark fiber and spectrum sharing, together with support for e-infrastructures, including media services like videoconferencing and lecture capturing & playback. More than 400 research & education institutions in the Nordics, with over 1.2 million users, are connected via the Nordic NREN networks, enabling scientists, educators,
    and students to work and share knowledge globally. NORDUnet is an active participant in the European NREN collaboration GÉANT and is a founding father of intercontinental NREN collaborations such as the Advanced North Atlantic (ANA) and Asia Pacific Europe Ring (AER) systems that are part of the Global Research and Education Network (GREN). In 2020, NORDUnet celebrated 40 years of Nordic NREN collaboration.

    About SURF

    SURF is the ICT cooperative of Dutch education and research institutions. The members, the owners of SURF, join forces to develop or procure the best possible digital services, work together on complex innovation issues and develop and share knowledge with each other.

    SURF actively collaborates with other European NRENs united in GÉANT and participates in global consortia like the Advanced North Atlantic (ANA) and Asia Pacific Europe Ring (AER).

    NetherLight, SURF’s Global Exchange Point (GXP) dedicated to research and education data in Amsterdam connects similar GXPs and advanced high-capacity networks for scientific and educational collaboration. The NetherLight GXP plays a major and vital role in the federation of research and education networks worldwide, also known as the Global Research and Education Network (GREN)

    About CANARIE

    CANARIE, together with its 13 provincial and territorial partners, forms Canada’s National Research and Education Network (NREN). This ultra-high-speed network connects Canada’s researchers and educators to each other and to global data, technology, and colleagues.

    To strengthen the security of Canada’s research and education sector, CANARIE collaborates with its partners in the NREN, government, academia, and the private sector to fund, implement, and support cybersecurity initiatives. CANARIE also provides identity management services to the academic community through eduroam and identity and access management services.

    Established in 1993, CANARIE is a non-profit corporation, with most of its funding provided by the Government of Canada.

  • International Collaboration at SuperComputing24: NetherLight/SURF participates in NICT-led experiment to advance high-speed data innovations

    International Collaboration at SuperComputing24: NetherLight/SURF participates in NICT-led experiment to advance high-speed data innovations

    During the SuperComputing 2024 (SC24) event in Atlanta, the Japanese National Institute of Information and Communications Technology (NICT) led an ambitious experiment using a global-scale experimental network. This network, established through collaboration with 19 international partners, connected Tokyo and Atlanta with 10 high-speed 100 Gbps paths, achieving a total capacity of 800 Gbps. The project showcased groundbreaking demonstrations of high-speed data transfer, anonymous communication, and innovative data management.

    One notable highlight was a data transfer experiment that reached 466 Gbps, and an award-winning anonymous communication demonstration by Osaka University achieved 588 Gbps while ensuring robust privacy. 

    NetherLight, the Global Exchange Point (GXP) run by SURF, played a critical role facilitating these complex experiments. Other key contributors included research and education networks and GXP’s from around the world, showcasing the collective effort required to achieve such innovation.

    This NICT experiment at SC24 emphasizes the importance of international collaboration and partnerships in global research and education networks. By working together in experiments like these, we can collectively show and assess the potential for transformative technologies in data handling and communication, essential for future scientific progress.

    For more details, pictures and graphics, and information about the contributing parties in this experiment, please read the full NICT press release here.

  • SC24: Advancing Distributed Hybrid Quantum Computing with SURF and NetherLight

    SC24: Advancing Distributed Hybrid Quantum Computing with SURF and NetherLight

    At SuperComputing 2024 (SC24) in Atlanta, an international collaboration showcased a groundbreaking demonstration of distributed hybrid quantum computing secured by advanced post-quantum cryptography (PQC) and quantum key distribution (QKD). This global effort brought together partners from Europe and the USA, highlighting how quantum and classical computing systems can be integrated and secured on a world scale.

    The Challenge and Opportunity of Quantum Computing

    Quantum computing holds immense potential for solving complex problems in fields like chemistry, biology, meteorology, and financial systems—challenges beyond the reach of classical computing. However, the technology’s cost, sensitivity, and limited availability present hurdles to its widespread application. Moreover, quantum computing threatens the security of current encryption systems, raising the stakes for robust, future-proof solutions.

    To address these challenges, the demonstration aimed to:

    1. Combine quantum computing with classical resources to improve accessibility and cost-effectiveness.

    2. Enable global distribution of these hybrid systems for broader researcher access.

    3. Protect these systems and data against threats in a post-quantum cryptographic environment.

    International Collaboration Driving Innovation

    This demonstration was the result of an international partnership involving European organizations (PSNC, GÉANT, SURF/NetherLight) and U.S. institutions (Internet2, ESnet, ICAIR/Northwestern University, StarLight). Together, they built a transatlantic hybrid quantum-classical computing network connecting testbeds in Poznan, Poland, and Atlanta, USA, using live production networking infrastructure.

    SURF and its NetherLight exchange played a pivotal role, enabling global connectivity alongside other major networks like GÉANT, Internet2, and SCinet. This collaborative approach leveraged expertise and resources from all partners to push the boundaries of what’s possible in quantum and classical computing integration.

    Technical Breakthroughs and Secure Data Transmission

    The demonstration showcased:

    – Hybrid quantum-classical computing integration using Quantum Processing Units (QPUs), CPUs, and GPUs.

    – High-speed data transmission over transatlantic links secured with PQC algorithms and QKD encryption.

    – Advanced security measures, including DWDM services for long-distance encryption and QKD technology for local network data security.

    This setup demonstrated the viability of a distributed quantum-classical infrastructure capable of supporting research use cases in fields like material science and optimization. By employing existing quantum computing systems with ~100 qubit capacity, the project advances the goal of achieving “quantum utility.”

    A Model for Future Innovation

    The SC24 demonstration underscores the power of international collaboration to solve complex challenges and drive technological breakthroughs. By integrating cutting-edge technologies and resources from diverse global partners, this project paves the way for the next generation of secure, distributed quantum computing infrastructure.

    SURF and NetherLight’s participation exemplifies their commitment to advancing science and innovation through global partnerships. Together with other partners, they are demonstrating how collective efforts can unlock the potential of quantum computing for research and education worldwide.

    The project was featured at SC24’s Network Research Exhibition, with a live presentation at the NRE Theatre, showcasing the transformative potential of distributed hybrid quantum computing.

    For further information and the full press release, please continue here.