SDPARA Cholesky 分解の性能 その２

以下の問題ですが、結局以下の計算時間と結果で終わりました。

phase.value = pdOPT
Iteration = 30
mu = +7.6198141164019624e-10
relative gap = +2.5334699957772818e-09
gap = +8.4617897755379090e-07
digits = +8.5962842347502679e+00
objValPrimal = -3.3399999916805575e+02
objValDual = -3.3400000001423473e+02
p.feas.error = +3.1984033556953551e-16
d.feas.error = +1.0913595360761974e-08
total time = 17850.511447

Time(sec) Ratio(% : MainLoop)
Predictor time = 17493.624322, 98.003089
Corrector time = 269.572116, 1.510202
Make bMat time = 2312.725925, 12.956394
Make bDia time = 0.000003, 0.000000
Make bF1 time = 0.000000, 0.000000
Make bF2 time = 0.000000, 0.000000
Make bF3 time = 0.000000, 0.000000
Make bPRE time = 0.000000, 0.000000
Make rMat time = 5.907221, 0.033094
Make gVec Mul = 1.917124, 0.010740
Make gVec time = 2.091932, 0.011719
copy gVec time = 0.000003, 0.000000
copy bMat time = 0.000000, 0.000000
symm bMat time = 1201.555535, 6.731376
Cholesky bMat = 13699.679463, 76.748585
Ste Pre time = 0.007532, 0.000042
Ste Cor time = 2.340430, 0.013112
solve = 528.789177, 2.962392
copy DyVec = 0.345439, 0.001935
sumDz = 1.070721, 0.005998
makedX = 10.572704, 0.059231
symmetriseDx = 0.211278, 0.001184
makedXdZ = 11.856249, 0.066421
xMatTime = 2.511153, 0.014068
zMatTime = 3.418216, 0.019150
invzMatTime = 0.000000, 0.000000
xMatzMatTime = 0.000000, 0.000000
EigxMatTime = 0.950597, 0.005325
EigzMatTime = 1.284709, 0.007197
EigxMatzMatTime = 0.000000, 0.000000
updateRes = 1.009352, 0.005655
EigTime = 2.235306, 0.012523
updateRes = 1.009352, 0.005655
EigTime = 2.235306, 0.012523
sub_total_bMat = 15537.348435, 87.043606
Main Loop = 17850.074360, 100.000000
File Check = 0.000000, 0.000000
File Change = 0.007599, 0.000043
File Read = 0.429488, 0.002406
File Trans = 3.602432, 0.020182
Total = 17850.511447, 100.002449

161989 = mDIM
1 = nBLOCK
1114 = bLOCKsTRUCT

◯計算サーバ：GPU 計算クラスタ
Intel Xeon + 4 GPU マシン（４台）
CPU：Xeon X5690(3.46GHz,6コア)×2
メモリ：192GB(16GB×12)
HDD：SATA500GB×2(システム、システムバックアップ)
NIC : GbE x 1 & Inifiniband(FDR) x 1
GPGPU：Tesla C2075(C2070)×4
OS：CentOS 6.3 for x86_64