/*
 *  -- clMAGMA (version 0.1) --
 *     Univ. of Tennessee, Knoxville
 *     Univ. of California, Berkeley
 *     Univ. of Colorado, Denver
 *     April 2012
 *
 **/

Performance results of the main routines provided in clMAGMA 0.1

Hardware
========
GPU : Tahiti  --- AMD Radeon HD 7900 
                  264 GB/s memory bandwidth, 3.79 TFLOPS SP, 947 GFLOPS DP
                  32 x 64 (2048 stream processors)
CPU : genesis --- 1 x six-core AMD Phenom II X6 1100T Processor @ 3.71 GHz

Software
========
GPU : AMD OpenCL BLAS v1.7 + 8.96 driver + clMAGMA 0.1 
CPU : MKL  11.1 + clMAGMA 0.1

//////////////////////////////////////////////////////////////////

genesis:~/cl_magma/testing> ./testing_spotrf_gpu 

  N    CPU GFlop/s (sec)    GPU GFlop/s (sec)    ||R_magma-R_lapack||_F / ||R_lapack||_F
========================================================================================
 1024      53.67 (  0.01)       8.34 (  0.04)         6.130442e-08
 2048      81.27 (  0.04)      77.71 (  0.04)         5.834552e-08
 3072      96.91 (  0.10)     168.37 (  0.06)         4.462960e-08
 4032     107.36 (  0.20)     198.68 (  0.11)         3.633282e-08
 5184     104.92 (  0.44)     275.32 (  0.17)         5.950510e-08
 6048     107.28 (  0.69)     396.13 (  0.19)         5.402312e-08
 7200     110.96 (  1.12)     485.88 (  0.26)         4.849893e-08
 8064     112.20 (  1.56)     545.95 (  0.32)         4.632914e-08
 8928     113.14 (  2.10)     604.08 (  0.39)         8.293850e-08
10560     114.17 (  3.44)     759.59 (  0.52)         7.422310e-08
12000     115.28 (  5.00)     784.68 (  0.73)         6.832436e-08
14080     116.42 (  7.99)     792.77 (  1.17)         5.892709e-08


genesis:~/cl_magma/testing> ./testing_sgeqrf_gpu

  M     N    CPU GFlop/s (sec)   GPU GFlop/s (sec)   ||R||_F / ||A||_F
======================================================================
 1024  1024    42.12 (  0.03)      20.81 (  0.07)       6.176354e-07
 2048  2048    90.82 (  0.13)      61.49 (  0.19)       6.405926e-07
 3072  3072   101.03 (  0.38)     145.76 (  0.27)       6.753147e-07
 4032  4032   108.98 (  0.80)     184.11 (  0.47)       9.110115e-07
 5184  5184   111.71 (  1.66)     261.13 (  0.71)       8.902191e-07
 6016  6016   114.25 (  2.54)     311.99 (  0.93)       8.948330e-07
 7040  7040   116.46 (  4.00)     368.41 (  1.26)       8.862651e-07
 8064  8064   118.03 (  5.92)     405.68 (  1.72)       8.980708e-07
 9088  9088   119.37 (  8.39)     457.57 (  2.19)       9.109222e-07
10176 10176   119.92 ( 11.72)     533.98 (  2.63)       9.485152e-07
12000 12000   122.29 ( 18.84)     588.51 (  3.92)       9.637155e-07
14080 14080   122.27 ( 30.44)     619.10 (  6.01)       9.287953e-07


genesis:~/cl_magma/testing>./testing_sgetrf_gpu                                                                                                                                                

  M     N    CPU GFlop/ (sec)s   GPU GFlop/s (sec)   ||PA-LU||/(||A||*N)
========================================================================
 1024  1024   12.16 (  0.06)      14.64 (  0.05)      1.366620e-09
 2048  2048   69.95 (  0.08)      48.85 (  0.12)      1.306808e-09
 3072  3072   91.21 (  0.21)      93.68 (  0.21)      1.272988e-09
 4032  4032  103.42 (  0.42)     199.98 (  0.22)      4.055000e-10
 4992  4992  109.52 (  0.76)     244.20 (  0.34)      3.598896e-10
 5952  5952  112.32 (  1.25)     283.39 (  0.50)      3.282320e-10
 7104  7104  115.31 (  2.07)     370.51 (  0.65)      2.999354e-10
 8064  8064  117.05 (  2.99)     436.46 (  0.80)      2.997826e-10
 9000  9000  118.73 (  4.09)     461.79 (  1.05)      1.829811e-09
10000 10000  120.55 (  5.53)     539.86 (  1.23)      2.010339e-09
12000 12000  120.32 (  9.57)     631.06 (  1.83)      2.295227e-09
14000 14000  121.72 ( 15.03)     691.03 (  2.65)      2.578583e-09

