Другие журналы
|
Lyfar'
77-30569/234879 Parallel algorithms for relational connection on graphics processors
Engineering Education # 10, October 2011 We present two parallel algorithms for relational connection on graphics processing units (GPU). The article considers the cases when the connection result is located in GPU memory and is returned by portions. Performance tests show five-fold advantage of GPU over CPU for non-indexed nested loops and 1.4-fold advantage for indexed nested loops.
|
|
|||||||||||||||||||||||||||||
|