| System Information | ||||
| Affiliation: | RIKEN Advanced Institute for Computational Scinece | URL: | http://www.aics.riken.jp/index_e.html | |
| Location: | Japan, Kobe, Hyogo | System Use: | Research | |
| System Manufacturer: | Fujitsu | System Name: | K computer | |
| Interconnect Manufacturer: | Fujitsu | Interconnect Type: | Tofu Interconnect | |
| Operating System: | Linux | MPI: | Parallelnavi Technical Computing Language V1.0L20 | |
| MPI Wtick: | 0.0000001 | BLAS: | Parallelnavi Technical Computing Language V1.0L20 | |
| Language: | C | Compiler: | Parallelnavi Technical Computing Language V1.0L20 | |
| Compiler Flags: | -Kfast,parallel,openmp,ocl -Xg (and "-mlcmain=main -SSL2BLAMP" for LINKFLAGS, "-Fwide" for Fortran files) | Processor Type: | Fujitsu SPARC64 VIIIfx | |
| Processor Speed: | 2 GHz | Total Processors: | 663552 | |
| Processors Entered: | 663552 | Processors determined: | 663552 | |
| Cores per chip: | 8 | HPL Processes: | 82944 | |
| MPI Processes: | 82944 | Threads Entered: | 8 | |
| Threads determined: | 8 | FLOPs per cycle: | ||
| Theoretical peak: | 10616.832 TFlop/s | Total memory: | 1327104 GiB | |
| FFT library: | FFTE 5.1beta | |||
| Explain Optimizations: | ||||
| G-HPL: adopted pipelined broadcast communication and promoted fine-grain overlapped data transfer G-RandomAccess: replaced MPI function calls with RDMA function calls, introduced 3-way data sorting and transfer to improve load imbalance, overlapped data transfer with data sorting, transformed data-sorting loop to use conditional move CPU instructions and applied loop distribution to table-updating loop S&EP-RandomAccess: applied loop distribution to table-updating loop and increased iterations of inner loops G-FFT: used the latest version of FFTE library EP-STREAM: inserted compiler directives to use a CPU feature of fast cache line allocation | ||||
| HPL | ||||
| HPL: | 9795.56 Tflop/s | HPL time: | 72072.9 | |
| HPL eps: | 1.11022e-16 | HPL Rnorm1: | 0.0000217012 | |
| HPL Anorm1: | 2550670 | HPL AnormI: | 2550600 | |
| HPL Xnorm1: | 59445100 | HPL XnormI: | 37.8587 | |
| HPL N: | 10192896 | HPL NB: | 448 | |
| HPL NProw: | 144 | HPL NPcol: | 576 | |
| HPL depth: | 1 | HPL NBdiv: | 2 | |
| HPL NBmin: | 8 | HPL CPfact: | L | |
| HPL CRfact: | L | HPL CPtop: | 0 | |
| HPL order: | C | |||
| HPL dMach EPS: | 1.110223e-16 | HPL sMach EPS: | 0.00000005960464 | |
| HPL dMach sfMin: | 2.225074e-308 | HPL sMach sfMin: | 1.175494e-38 | |
| HPL dMach Base: | 2 | HPL sMach Base: | 2 | |
| HPL dMach Prec: | 2.220446e-16 | HPL sMach Prec: | 0.0000001192093 | |
| HPL dMach mLen: | 53 | HPL sMach mLen: | 24 | |
| HPL dMach Rnd: | 1 | HPL sMach Rnd: | 1 | |
| HPL dMach eMin: | -1021 | HPL sMach eMin: | -125 | |
| HPL dMach rMin: | 2.225074e-308 | HPL sMach rMin: | 1.175494e-38 | |
| HPL dMach eMax: | 1024 | HPL sMach eMax: | 128 | |
| HPL dMach rMax: | 1.797693e308 | HPL sMach rMax: | 3.402823e38 | |
| dweps: | 1.110223e-16 | sweps: | 0.00000005960464 | |
| PTRANS | ||||
| PTRANS: | 16551.8 GB/s | PTRANS time: | 12.554 seconds | |
| PTRANS residual: | 0 | PTRANS N: | 5096448 | |
| PTRANS NB: | 193 | PTRANS NProw: | 144 | |
| PTRANS NPcol: | 576 | |||
| STREAM | ||||
| S-STREAM Copy: | 45.1667 GB/s | S-STREAM Scale: | 45.1663 GB/s | |
| S-STREAM Add: | 43.7775 GB/s | S-STREAM Triad: | 46.5233 GB/s | |
| EP-STREAM Copy: | 44.6461 GB/s | EP-STREAM Scale: | 44.5609 GB/s | |
| EP-STREAM Add: | 43.7563 GB/s | EP-STREAM Triad: | 46.5051 GB/s | |
| STREAM Vector Size: | 417531221 | STREAM Threads: | 8 | |
| RandomAccess | ||||
| S-RandomAccess: | 0.107065 Gup/s | EP-RandomAccess: | 0.106791 Gup/s | |
| G-RandomAccess: | 471.939 Gup/s | G-RandomAccess N: | 70368744177664 | |
| G-RandomAccess time: | 596.423 seconds | G-RandomAccess Check Time: | 817.33 seconds | |
| G-RandomAccess Errors: | 2097152 | G-RandomAccess Errors Fraction: | 0.0000000298023 | |
| G-RandomAccess TimeBound: | -1 | G-RandomAccess ExeUpdates: | 281474976710656 | |
| RandomAccess N: | 1073741824 | |||
| FFT | ||||
| S-FFT: | 6.88977 GFlop/s | EP-FFT: | 6.88461 GFlop/s | |
| MPIFFT: | 205936 GFlop/s | MPIFFT N: | 12899450880000 | |
| MPIFFT Max Error: | 0.00000000000000380413 | MPIFFT time0: | 0.000003402 seconds | |
| MPIFFT time1: | 2.94935 seconds | MPIFFT time2: | 1.79649 seconds | |
| MPIFFT time3: | 2.76618 seconds | MPIFFT time4: | 2.95597 seconds | |
| MPIFFT time5: | 3.00096 seconds | MPIFFT time6: | 0.0000017999 seconds | |
| FFTEnblk: | 16 | FFTEnp: | 8 | |
| FFTEl2size: | 6291456 | |||
| DGEMM | ||||
| S-DGEMM: | 116.339 GFlop/s | EP-DGEMM: | 99.2725 GFlop/s | |
| DGEMM N: | 20433 | |||
| RandomRing Latency/Bandwidth | ||||
| RandomRing Latency: | 6.2523 usec | RandomRing Bandwidth: | 0.187285 GB/s | |
| NaturalRing Latency/Bandwidth | ||||
| NaturalRing Latency: | 4.98281 usec | NaturalRing Bandwidth: | 3.11904 GB/s | |
| PingPong Latency/Bandwidth | ||||
| Maximum PingPong Latency: | 7.44663 usec | Maximum PingPong Bandwidth: | 4.51146 GB/s | |
| Minimum PingPong Latency: | 2.95162 usec | Minimum PingPong Bandwidth: | 4.35193 GB/s | |
| Average PingPong Latency: | 4.79643 usec | Average PingPong Bandwidth: | 4.44832 GB/s | |
| Size of Data Types | ||||
| char: | 1 byte | short: | 2 bytes | |
| int: | 4 bytes | long: | 8 bytes | |
| void ptr: | 8 bytes | float: | 4 bytes | |
| double: | 8 bytes | size t: | 8 bytes | |
| s64Int: | 8 bytes | u64Int: | 8 bytes | |
| OpenMP | ||||
| M OpenMP: | 200805 | OpenMP Num Threads: | 8 | |
| OpenMP Num Procs: | 8 | OpenMP Max Threads: | 8 | |
| Memory | ||||
| MemProc: | -1 | MemSpec: | -1 | |
| MemVal: | -1 | |||
| CPS | ||||
| CPS_HPCC_FFT_235: | 1 | CPS_HPCC_FFTW_ESTIMATE: | 0 | |
| CPS_HPCC_MEMALLCTR: | 0 | CPS_HPL_USE_GETPROCESSTIMES: | 0 | |
| CPS_RA_SANDIA_NOPT: | 1 | CPS_RA_SANDIA_OPT2: | 0 | |