Changes between Version 5 and Version 6 of WorkingGroups/HPC
- Timestamp:
- 2014-11-05T22:49:20+01:00 (10 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
WorkingGroups/HPC
v5 v6 30 30 31 31 == Some ideas...:[[BR]] == 32 A strong improvement of NEMO scalability is needed to be able to take advantage of the new machines. This means a deep review/rewrite of NEMO code at some point in the futur (beyond 5 years). At the same time, we already know that CMIP7 won't use an ocean model that has not been strongly tested and validated and will stick to a NEMO model close tothe existing one. [[BR]]32 A strong improvement of NEMO scalability is needed to be able to take advantage of the new machines. This probably means a deep review/rewrite of NEMO code at some point in the futur (beyond 5 years from now?). At the same time, we already know that CMIP7 won't use an ocean model that has not been strongly tested and validated and will stick to a NEMO model not so far from the existing one. [[BR]] 33 33 This means that we need to: 34 34 1) keep improving the current structure of NEMO so it works quite efficiently for almost 10 more years (until the end of CMPI7). [[BR]] 35 2) start to work on a new structure that would be tested from year 5 and fully tested and validatedfor CMIP8 in about 10 years. [[BR]]35 2) start to work on a new structure that would fully tested and validated at least for CMIP8 in about 10 years. [[BR]] 36 36 37 Based on this, we propose to divide the work according to 3 temporal window [[BR]]37 Based on this, we propose to divide the work according to 3 temporal windows [[BR]] 38 38 39 '''0-3 years''': improvements with existing code: Do basic optimisation work: [[BR]] 40 1) reduce the number of communications: remove useless communications (a lot) [[BR]] 41 2) reduce the number of communications: do less and bigger communications (group communications, use large halo) [[BR]] 39 '''0-3 years''': improvements with existing code: [[BR]] 40 0) remove solvers (to be done in 3.7) 41 1) reduce the number of communications: do less and bigger communications (group communications, use larger halo). main priority: communications in the time splitting. [[BR]] 42 2) reduce the number of communications: remove useless communications (a lot) [[BR]] 42 43 3) introduce asynchronous communications 43 44 3) check code vectorization (SIMD instructions) [[BR]] … … 45 46 '''0-5 years''': improvements through the introduction of OpenMP: [[BR]] 46 47 work initialed by CMCC. 47 "Basic" implementation such as tiling may be efficient with many cores processors? 48 test different way to find new sources of parallelism. 49 OpenMP4, OpenACC? 48 implementation such as tiling may be efficient with many cores processors? review lbclnk to be able to deal with MPI and OpenMP 49 OpenMP along the vertical axis? Find a way to remove implicit schemes? 50 test different way to find new sources of parallelism for example with the help of OpenMP4 51 test OpenACC (not that far from OpenMP)? 50 52 51 53 '''beyond 5 years''': [[BR]]