Oct. 28 — On October 30, 2014, the Research Organization for Information Science and Technology (RIST) and the Partnership for Advanced Computing in Europe (PRACE) will sign a memorandum of understanding (hereinafter called MOU) on information exchange concerning the promotion of shared use of supercomputers. This will help RIST and PRACE to perform their projects more effectively and efficiently.
RIST is promoting the shared use of supercomputer K and the other HPCI computer systems in Japan, and PRACE is promoting the shared use of top level (Tier-0) supercomputers of the six sites in Europe.
As it is extremely beneficial for both parties to exchange information in the area of promotion of shared use of supercomputers, a MOU with following contents is concluded;
- Exchange of information; Mutual exchange of experiences and knowledge in user selection and user supports etc. is helpful for both parties in order to execute their projects more effectively and efficiently.
- Holding technical meetings; Technical meetings are appropriately held to perform information exchange more effectively.
On October 30, Sergi Girona, Managing Director of PRACE, will visit the RIST Kobe Center for the signing of the MOU and the first information exchange meeting between both parties will be held there. He will also join the First Project Report Meeting of the HPCI System including the K computer, hosted by RIST, to be held in Tokyo on October 31, 2014. He will give a special talk on the present status of the use of supercomputers and future prospects in Europe.
In the European Union, the following HPC ecosystem is in place:
- Tier-0 EU level, to which access is provided by PRACE
- Tier-1 national level to which access is provided via national calls and the DECI programme for international exchange
- Tier-2 regional center/university level
PRACE provides the highest hierarchical level of Tier-0, composed of six sites. It provided >15 Petaflops at 6 sites in 2013. Development is underway with the goal to reach 19 Petaflops in 2014. PRACE 2.0 will start from 2015, aiming to provide computing resources of over 50 Petaflops by 2020.
—
Source: PRACE