Ant

LAPACK has been designed to supersede LINPACK [38] and EISPACKDense and band matrices are provided for, but not general sparse matrices9) What about shared memory machines? Are there multi

Degrees of LatencyThe latency of data access becomes greater with each cache level. Latency of memory access is best measured in CPU clock cycles. One cycle occupies from 4 to 6 nanoseconds, dep

来自：http://www.netlib.org/utk/people/JackDongarra/la-sw.htmlFreely Available Software for Linear Algebra (May 2013)             

N= 500float  no pivoting: 23:19:13:~/..er/inverse$ ./main 500*****************************************************// floatno pivoting:no_pivot_inverse:0.088006 secondsno_pivot_mult:0.14800

23:37:01:~/..er/inverse$ ./main 2000*****************************************************no pivoting:no_pivot_inverse:10.924682 secondsno_pivot_mult:67.480217 secondsA_original*Ainv:     

poisson5point1000＊1000  cg: 1021 收敛gmres(50)   261gmres(100) 217cg_m  1269  时间都差不多bicgstab   无穷大，不收敛