Hello and help with cusparse cusparse

[cletus_42]

Hello all, I am VERY glad this forum is out there.

I want to build a set of sparse solver(s) for a program and found the cusparse library and it looks very like a good place to start to create the “building blocks” of my own sparse solver(s). Suppose I have a sparse matrix that has already been compressed using the Compressed Sparse Row format and want to use the cusparse sparse matrix-vector multiplication call? Can anyone point me to a simple example of this? I find a lot of information about very involved program(s), but I just want something very simple (if possible) in this area.

For example:
let ‘val’ be the non-zero value vector (derived from original sparse 2D matrix),
‘ia’ be the row pointer(s), and ‘ja’ be the column indices. I would like to store
the result of the CSR spMV operation from cusparse in another vector, ‘x’.

Much thanks for any help.

[cletus_42]

Hello all, I am VERY glad this forum is out there.

I want to build a set of sparse solver(s) for a program and found the cusparse library and it looks very like a good place to start to create the “building blocks” of my own sparse solver(s). Suppose I have a sparse matrix that has already been compressed using the Compressed Sparse Row format and want to use the cusparse sparse matrix-vector multiplication call? Can anyone point me to a simple example of this? I find a lot of information about very involved program(s), but I just want something very simple (if possible) in this area.

For example:
let ‘val’ be the non-zero value vector (derived from original sparse 2D matrix),
‘ia’ be the row pointer(s), and ‘ja’ be the column indices. I would like to store
the result of the CSR spMV operation from cusparse in another vector, ‘x’.

Much thanks for any help.

[cletus_42]

Hello all, I am VERY glad this forum is out there.

I want to build a set of sparse solver(s) for a program and found the cusparse library and it looks very like a good place to start to create the “building blocks” of my own sparse solver(s). Suppose I have a sparse matrix that has already been compressed using the Compressed Sparse Row format and want to use the cusparse sparse matrix-vector multiplication call? Can anyone point me to a simple example of this? I find a lot of information about very involved program(s), but I just want something very simple (if possible) in this area.

For example:

let 'val' be the non-zero value vector (derived from original sparse 2D matrix),

'ia' be the row pointer(s), and 'ja' be the column indices. I would like to store

the result of the CSR spMV operation from cusparse in another vector, 'x'.

Much thanks for any help.

Never mind I figured it out after digging into the library documents on CUDA.