Two hundred years ago, a young man penned the first note on paper, and the king of the piano, Liszt, praised it with "We can never surpass or match it."

His name was Chopin, and what he wrote was called "Nocturne."

Today, in the digital age, when the sound of the piano by the Vistula River rings out again, standing on the land of Poland, the birthplace of the master Chopin, a romantic digital nocturne is also slowly unfolding.

Polish National Supercomputing Center PCSS:

A Digital Aria under the Starry Sky

In the long river of human music history, Chopin is called a master because he broke free from the shackles of classicalism in his creative style and opened up a whole new era.

The Supercomputing and Networking Center PCSS (Poznańskie Centrum Superkomputerowo-Sieciowe), affiliated with the Polish Academy of Sciences, is also working with Huawei OceanStor Pacific distributed storage to explore and innovate digitally, breaking through the entire night sky of European scientific research.

PCSS is currently the most advanced supercomputing center in Poland and even in Europe. By building and integrating many powerful computing cluster resources, PCSS provides the most advanced high-performance computing (HPC) services for nearly 1000 research institutions and professional schools in Europe, to solve larger-scale and more complex scientific problems in various industries.

Advertisement

Among them, radio astronomy is the most wonderful and stunning application.Humans tirelessly cast their gaze into the deep cosmos, thirsting to understand the universe that has never been fully revealed. Nourished by astrophysics and radio wave theory, radio astronomy was born.

Unlike ordinary light beams, the electromagnetic waves emitted from celestial bodies, with their lower frequencies and longer wavelengths, can shield the effects of adverse weather, bypass the obstruction of dust and nebulae, and ultimately be collected, converged, and recorded by the towering radio telescopes on the ground, restoring more accurate images and coordinates.

Those dark corners invisible to the naked eye burn brightly in the spectrum of electromagnetic waves. In the vast universe, radio telescopes can sense every object and capture every message. All human fantasies and confusions about pulsars, black holes, and the Big Bang are unraveled by these colossal "cameras."

In the 1990s, a distributed radio telescope project named LOFAR (Low-Frequency Array) was initiated in Europe. After multiple iterations, this has become one of the largest and most sensitive low-frequency radio telescope systems on Earth, and is now one of Europe's most powerful tools for research in physics and astronomy.

Among them, three LOFAR stations were established in Poland, and the task of the Polish PCSS supercomputing center is to receive signals from tens of thousands of dipole antennas of LOFAR. Then, by deploying the most advanced interferometric measurement algorithms, they stitch, calibrate, and process the raw data to trace the long-awaited secrets from billions of light-years away.

As numbers flow, mysteries are quietly unlocked, and the grand cosmic epic unfolds before our eyes.

Seeing the Universe in the "Lake":

Connecting the Most Distant Unknowns

In 2022, European scientists used LOFAR to process 3,500 hours of observational data, creating the most accurate and largest radio map of the universe, revealing approximately one million previously undiscovered celestial bodies and 25,000 black holes to the world. The data from this single experiment alone occupied 8 petabytes of storage space, equivalent to the capacity of 20,000 laptop computers.Scientists Utilize LOFAR to Output Authentic Imagery Details from Standard to Highest Resolution

The unprecedentedly powerful LOFAR array can provide "super images" with a resolution 20 times superior to common standards, revealing unique details and clues in the vast universe with a precision two orders of magnitude higher than industry norms. Correspondingly, the massive amount of data generated by LOFAR each year far exceeds initial estimates.

According to statistics, the accumulated data volume in LOFAR's history has exceeded 55 petabytes (PB) and continues to rise at a rate of 5PB annually, even producing 2.5 gigabytes (GB) of raw data every second.

This largest-ever astronomical data collection, a part of which is stored within the PCSS supercomputing center in Poland, is not only a concern in the field of radio astronomy. Rare massive scientific data from fields such as genomic sequencing and ultraviolet satellites have become the primary headache for PCSS. Under the "data-driven" development strategy, PCSS has decided to take the lead in constructing the National Data Storage (NDS) project to create a scientific data lake that seamlessly integrates with high-performance computing, big data, and AI, among other emerging technologies.

In PCSS's vision, there is a need for a distributed "super data lake" with a capacity exceeding 1000PB—featuring a flexible and open architecture that connects several supercomputing nodes across Poland, pooling all data into the lake for immediate application.

During the business preparation phase before computation, massive data from various disciplines must go through steps such as collection, cleaning, condensation, publication, and archiving, which requires storage that can "hold a lot": equipped with diverse protocol interfaces and nearly unlimited scalability. In the data analysis phase, each computation involves reading and writing data sets of hundreds of GB, which demands storage that "runs fast": capable of matching high-performance computing, big data, AI, and other mixed loads, with a throughput capacity of at least 50GB/s to meet experimental schedules.

PCSS began seeking assistance from professional storage manufacturers worldwide. In 2023, PCSS deployed and tested a 48PB Huawei OceanStor Pacific distributed storage system across nine sites. It features industry-unique file/object/big data multi-protocol lossless intercommunication, as well as high-performance distributed parallel clients (DPC), demonstrating unimaginable performance and device power consumption, standing out after rigorous selection.

Notably, in a competition with the Ceph+ storage server solution, OceanStor Pacific distributed storage, with its industry-leading high-density hardware design combined with the intelligent tiering capability of SSD and HDD models, reduced the equipment room space by 65% and lowered the comprehensive electricity cost by two-thirds over five years, ultimately becoming PCSS's top choice.

"Huawei's solution has significantly improved the data infrastructure in our scientific HPC systems," said Norbert Meyer, Director of HPC and Data Infrastructure Business at PCSS, "We were very pleasantly surprised by the stability and reliability demonstrated by the storage system during testing."Such a "super data lake" has become a digital support under the cutting-edge technology of the Polish PCSS supercomputing center, secretive yet great.

Huawei's OceanStor Pacific distributed storage, just like its name, embraces data with the momentum of an ocean, carrying the grand and magnificent imagination of countless scientific researchers.

Just like the nocturnes composed by Chopin, at first listen they seem to be tender sighs, but in reality, they contain a surging and passionate force from the depths of the soul.

For the Polish PCSS supercomputing center, behind the continuous renewal of data, advanced storage capabilities are accumulating energy, becoming the confidence to look up to the heavens, and endowing PCSS with a new starting point to explore the vast world. With Huawei's help, PCSS's storage system has entered the top ten of the world's most powerful HPC storage rankings, the IO500.

Nowadays, industry insiders often use "calculating the sky, the sea, the earth, and people" to describe the ubiquitous application scenarios of high-performance computing. Huawei's OceanStor Pacific distributed storage, which has been deeply involved in the HPC field for many years, is also striving to move forward, illuminating the vast galaxy of thousands of industries with practitioners in production, academia, and research.

Believe that more stories have already set sail among the mountains and rivers.