BSGP: bulk-synchronous GPU programming

BSGP: bulk-synchronous GPU programming,10.1145/1360612.1360618,ACM Transactions on Graphics,Qiming Hou,Kun Zhou,Baining Guo

BSGP: bulk-synchronous GPU programming   (Citations: 20)
BibTex | RIS | RefWorks Download
We present BSGP, a new programming language for general pur- pose computation on the GPU. A BSGP program looks much the same as a sequential C program. Programmers only need to sup- ply a bare minimum of extra information to describe parallel pro- cessing on GPUs. As a result, BSGP programs are easy to read, write, and maintain. Moreover, the ease of programming does not come at the cost of performance. A well-designed BSGP compiler converts BSGP programs to kernels and combines them using opti- mally allocated temporary streams. In our benchmark, BSGP pro- grams achieve similar or better performance than well-optimized CUDA programs, while the source code complexity and program- ming time are significantly reduced. To test BSGP's code efficiency and ease of programming, we implemented a variety of GPU appli- cations, including a highly sophisticated X3D parser that would be extremely difficult to develop with existing GPU programming lan- guages.
Journal: ACM Transactions on Graphics - TOG , vol. 27, no. 3, 2008
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
Sort by: