噓~

libmkl.so  for dynamic
librml.so  for dynamic w VML/VSL
 
-lguide  使連結threading library libguide(使support multi-threading)

mkl 與cmkl (cluster mkl)有什麼差別?
 1. linking in commend line 不同
 2. 對於BLAS, LAPACK 等是一樣的 但CMKL多了scalable LAPACK 
   舉例來說:    mkl 與許你解一個矩陣的eigenvalues 

linker example :
http://software.intel.com/sites/products/documentation/hpc/mkl/lin/MKL_UG_linking_your_application/Linking_Examples.htm
 
static linking:

mkl 使用手冊
(all)                       http://www.intel.com/cd/software/products/apac/zho/358630.htm  
(windows 版本)  http://softwarecommunity.intel.com/isn/downloads/softwareproducts/pdfs/347468.pdf
(linux :information )http://softwarecommunity.intel.com/isn/downloads/softwareproducts/pdfs/347460.pdf
8.0(

?yyzzz : ? would be  s,d,c,z
s ,real,*4
d,real,*8
c,complex,*8
z,complex,*16
------------------------------------------------------------------------------------------------
yy could be
--------------
In BLAS level 2 and 3,
ge general matrix
gb general band matrix
sy symmetric matrix
sp symmetric matrix (packed storage)
sb symmetric band matrix
he Hermitian matrix
hp Hermitian matrix (packed storage)
hb Hermitian band matrix
tr triangular matrix
tp triangular matrix (packed storage)
tb triangular band matrix.
ge general
gb general band
gt general tridiagonal
po symmetric or Hermitian positive-definite
pp symmetric or Hermitian positive-definite (packed storage)
pb symmetric or Hermitian positive-definite band
pt symmetric or Hermitian positive-definite tridiagonal
sy symmetric indefinite
sp symmetric indefinite (packed storage)
he Hermitian indefinite
hp Hermitian indefinite (packed storage)
tr triangular
tp triangular (packed storage)
tb triangular band
--------------------
BLAS level 1
c conjugated vector
u unconjugated vector
g Givens rotation
-------------------
BLAS level 2
mv matrix-vector product
sv solving a system of linear equations with matrix-vector operations
r rank-1 update of a matrix
r2 rank-2 update of a matrix.
------------------
BLAS level 3
mm matrix-matrix product
sm solving a system of linear equations with matrix-matrix operations
rk rank-k update of a matrix
r2k rank-2k update of a matrix
------------------------------------------------------------------------------------------------
-------------------
The examples below illustrate how to interpret BLAS routine names:
ddot <d> <dot>: double-precision real vector-vector dot product
60
2 Intel® Math Kernel Library Reference Manual
cdotc <c> <dot> <c>: complex vector-vector dot product, conjugated
<sc> <asum>: sum of magnitudes of vector elements, single precision
real output and single precision complex input
scasum
cdotu <c> <dot> <u>: vector-vector dot product, unconjugated, complex
sgemv <s> <ge> <mv>: matrix-vector product, general matrix, single precision
<z> <tr> <mm>: matrix-matrix product, triangular matrix,
double-precision complex.
--------------------------------
------------------------------------------------------------------------------------------------
zzz could be
---------------------
BLAS Level3
 trf  form a triangular matrix factorization
3 Intel® Math Kernel Library Reference Manual
trs solve the linear system with a factored matrix
con estimate the matrix condition number
rfs refine the solution and compute error bounds
tri compute the inverse matrix using the factorization
equ equilibrate a matrix.
------------------------
For example, the sgetrf routine performs the triangular factorization of general real matrices
in single precision; the corresponding routine for complex matrices is cgetrf.
For driver routines, the names can end with -sv (meaning a simple driver), or with -svx (meaning an expert driver).

Names of the LAPACK computational and driver routines for Fortran 95 interface in Intel MKL  are the same as Fortran 77 names but without the first letter that indicates the data type. For example, the name of the routine that performs triangular factorization of general real matrices in Fortran 95 interface is getrf. Different data types are handled through defining a specific internal parameter that refers to a module block with named constants for single and double precision.