Thousands of super mico computers were supplied to CERN to help advance HPC research

CERN (the European Organization for Nuclear Research) brings together more than 17,000 experts from across the globe to study the tiniest particles in the cosmos. Some of the largest and most intricate scientific equipment is used to comprehend nature's underlying structure better and expand human understanding.

Several CERN discoveries have had a significant influence on a variety of fields, ranging from medicine to computer science. As a long-standing supplier to CERN, Supermicro offers corporate computing technology that facilitates cutting-edge particle physics research. As the quest to comprehend the beginnings of the cosmos develops, new computational resources are continuously examined and updated.

The relationship between CERN and Supermicro facilitates the incorporation of new technologies into the high-performance computing cluster that handles Large Hadron Collider data (LHC). Scientists at CERN continue to investigate enterprise innovations for CERN in partnership with other worldwide technology producers.

Challenges

Using particle accelerators, CERN is continually at the forefront of research to discover the composition of the cosmos. To comprehend the findings of the trials, however, a substantial amount of processing power is necessary. Hence, it must be brought up to date with the most recent improvements in server, storage, and networking technologies while keeping an eye on the overall power usage of the computer system.

Solution

Before any purchase, the infrastructure technical team at CERN labored for months to reach the market for the most recent technological advancements. Although time-consuming, it is a qualifying procedure that is essential to maintaining CERN at the vanguard. More critically, it enables cooperation with relevant partners to evaluate options in terms of performance, cost, operating expenses, power, density, etc, and so on.

After a protracted procedure, CERN chose the Supermicro AMD BigTwin A+ 2124BT-HNTR server. So, CERN bought more than 900 Supermicro BigTwin systems, each with four nodes, for around 3,600 server nodes. Scientists at CERN execute several simulations and tasks. Most are used for batch computing tasks associated with physics-event reconstruction, data analysis, and simulation. In addition, some of them are employed as disk storage front-ends to gather physics data from experiments.

Supermicro BigTwin servers are intelligent and cost-effective investments for businesses and institutions constructing, expanding, or future-proofing sophisticated computing infrastructures. The architecture of Supermicro BigTwin is the basis for the most sophisticated server systems in HPC. This high-performance, high-density systems have optimal airflow for energy-efficient cooling, simple maintenance, and high availability thanks to hot-swappable nodes and redundant power supply modules.

Benefits

The performance of several simulation programs at CERN improved. Due to the shared cooling and power systems, the high density of Supermicro BigTwin systems with AMD EPYC CPUs resulted in lower power consumption than anticipated. With the AMD EPYCTM 7XXX CPU, programs ran quicker than prior generations of CPUs, enabling scientists to do more research and better understand our planet.

"Supermicro's dedication to the design and engineering of the Supermicro BigTwin architecture has steadily grown over many years of collaboration. Thousands of compute nodes based on the most recent AMD EPYC-based Supermicro BigTwin platform were implemented at the CERN computing facilities in 2021. This indicates the performance and dependability of these systems' superior quality."

Download Supermicro's whitepaper to learn more about cloud computing servers only on Whitepapers Online.