Font Size: a A A

Design And Implementation Of Parallel Encrypt And Decrypt System Based On GPU

Posted on:2014-04-02Degree:MasterType:Thesis
Country:ChinaCandidate:L H ZhouFull Text:PDF
GTID:2268330422951985Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of Internet, data security becomes more and moresignificant. Data encryption is an important means to protect the data security. AESencryption algorithm is one of the important algorithms to encrypt data. However,when the data is large, the traditional AES encryption algorithm is slow.With the development of graphics processing Unit, the computing ability ofgraphics processor is far higher than that of CPU in same period. In large-scale dataparallel computing, Graphics processor plays an increasingly important role.Compute Unified Device Architecture, especially the parallel architecture of OpenCompute Language appears, which simplifies programming the graphics processor.It can obviously accelerate encrypt the large data using graphics processor toimprove the AES encryption algorithm. This topic comes from Americantechnology co., LTD. And the main research contents concludes utilizing the highparallel computing ability of graphics processor to accelerate the AES algorithm,speeding up the encryption and saving time when encrypt the large data. Use OpenCL programming heterogeneous architecture of AES algorithm working mode onthe graphics processor to realize the parallel acceleration; use Open CLsynchronization mechanism to complete the synchronization among multithreading;Use double thread technique improved the utilization of graphics processor andCPU; Use the memory object model of Open CL to optimize the system locally. Atlast, it uses the Parallel Performance analysis tools AMD APP Profiler and theParallel Performance Analyzer (PPA) to test and analyze the time performance ofthe system.
Keywords/Search Tags:Graphics Processing Unit, OpenCL, parallel computing, multi-thread, encrypt and decrypt data
PDF Full Text Request
Related items