Description of the RI

During recent years, applications of latest computational techniques and technologies generated breakthrough in various research fields. Traditionally, there is a strong demand for cutting-edge IT-solutions in particle, astroparticle, nuclear and theoretical physics, and related data analysis. KFKI at Csillebérc pioneered data collections in experiments and their analysis from early times. This knowledge was inherited by many research institutes of the campus including KFKI RMKI, by providing and hosting novel IT-technologies. Since 2000 large scale and high performance computing clusters were installed, and from 2009 GPU-technology and multi-platform, hybrid architectures were applied in different research fields.

Today a state-of-the-art research and development direction is the application of FPGA-based computers for emulation of quantum computers, to pave the way for quantum machine learning and real quantum computer technologies of near-future.

Recently the applied-science motivated Wigner Scientific Computing Laboratory (WSCLAB) has been established, integrating the decade-old Wigner GPU Laboratory with the existing large-scale multiprocessor CPU clusters. This WSCLAB is located in the Wigner Datacenter (1 MW IT-power, 100 racks, and 10 Gb/s internet connections). The Mission of the WSCLAB is serving as a knowledge center, providing computational capacity and technology for research projects and scientists. Tutorial events and workshops have been organized: GPU Day series at 12 times, Academy-Industry Matching Event on IT-solutions since 5 years.

Wigner Scientific Computing Laboratory hosts the following units and technologies:

• CERN ALICE/CMS Tier2 Cluster [4000 vCPU, 10 TB RAM, 3.6 PB storage]
• CERN ALICE Analysis Facility [3700 vCPU, 7.5 TB RAM, 2.3 PB storage]
• VIRGO/ET Tier2 Cluster [1000 vCPU, 3 TB RAM, 150 TB storage]
• GPU Units [8x Tesla A100 SXM4 and 14x Tesla T4]
• Maxeler FPGA Unit as Quantum Computing Simulator

Description of the RI

Wigner Scientific Computing Laboratory (WSCLAB) hosts dedicated Tier2 clusters for CERN and VIRGO/ET Collaborations serving their computing requests. In parallel, knowledge transfer is offered for interested scientists, who are responsible for large scale, high-capacity clusters in other collaborations.

In the ALICE AF unit research projects are optimized for effective usage of multi-core processors in large scale IT-systems to reach maximal performance of the hardware elements.

In the GPU unit the users are able to test and develop their application on special architectures, containing different GPU-hardware, and they receive support on application of GPU-based languages.

In the FPGA unit a quantum computer emulator has been established on the basis of DataFlow technology. This can be used for study of different aspects of quantum programming and quantum machine learning protocols.

The mission of WSCLAB is serving IT-expert researchers with novel IT-technologies, focusing on applications in research and industrial developments. The existing local background is based on two decades activity in the frontier of IT-developments and this guarantees successful cooperation and knowledge transfer. Using the resources of WSCLAB could speed up research activity of experienced users, and serve an opportunity for beginners to follow recent IT-developments in the fields of artificial intelligence, machine learning and quantum programming. Seminars and workshop helps to increase the effectivity of knowledge transfer.