GDL

Effective Sparse Matrix Representation For The GPU Architectures

ISSN : 2230 - 9616
Undergraduate Theses from gdlhub / 2017-08-14 11:52:34
By : B.Neelima and Prakash S.Raghavendra, International Journal of Computer Science, Engineering and Applications
Created : 2012-07-02, with 1 files

Keyword : GPU, CPU, SPMV, CSR, COO, CSR-vector
Subject : Effective Sparse Matrix Representation For The GPU Architectures
Url : http://airccse.org/journal/ijcsea/papers/2212ijcsea13.pdf
Document Source : Internet

General purpose computation on graphics processing unit (GPU) is prominent in the high performance


computing era of this time. Porting or accelerating the data parallel applications onto GPU gives the


default performance improvement because of the increased computational units. Better performances can


be seen if application specific fine tuning is done with respect to the architecture under consideration. One


such very widely used computation intensive kernel is sparse matrix vector multiplication (SPMV) in sparse


matrix based applications. Most of the existing data format representations of sparse matrix are developed


with respect to the central processing unit (CPU) or multi cores. This paper gives a new format for sparse


matrix representation with respect to graphics processor architecture that can give 2x to 5x performance


improvement compared to CSR (compressed row format), 2x to 54x performance improvement with respect


to COO (coordinate format) and 3x to 10 x improvement compared to CSR vector format for the class of


application that fit for the proposed new format. It also gives 10% to 133% improvements in memory


transfer (of only access information of sparse matrix) between CPU and GPU. This paper gives the details


of the new format and its requirement with complete experimentation details and results of comparison.

Description Alternative :

General purpose computation on graphics processing unit (GPU) is prominent in the high performance


computing era of this time. Porting or accelerating the data parallel applications onto GPU gives the


default performance improvement because of the increased computational units. Better performances can


be seen if application specific fine tuning is done with respect to the architecture under consideration. One


such very widely used computation intensive kernel is sparse matrix vector multiplication (SPMV) in sparse


matrix based applications. Most of the existing data format representations of sparse matrix are developed


with respect to the central processing unit (CPU) or multi cores. This paper gives a new format for sparse


matrix representation with respect to graphics processor architecture that can give 2x to 5x performance


improvement compared to CSR (compressed row format), 2x to 54x performance improvement with respect


to COO (coordinate format) and 3x to 10 x improvement compared to CSR vector format for the class of


application that fit for the proposed new format. It also gives 10% to 133% improvements in memory


transfer (of only access information of sparse matrix) between CPU and GPU. This paper gives the details


of the new format and its requirement with complete experimentation details and results of comparison.

Give Comment ?#(0) | Bookmark

Property	Value
Publisher ID	gdlhub
Organization	I
Contact Name	Herti Yani, S.Kom
Address	Jln. Jenderal Sudirman
City	Jambi
Region	Jambi
Country	Indonesia
Phone	0741-35095
Fax	0741-35093
Administrator E-mail	elibrarystikom@gmail.com
CKO E-mail	elibrarystikom@gmail.com