• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ÇÐȸÁö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ÇÐȸÁö > µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

µ¥ÀÌÅͺ£À̽º ¿¬±¸È¸Áö(SIGDB)

Current Result Document : 4 / 13 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) GPGPU ±â¹Ý Á¶ÀÎ ¿¬»ê º´·ÄÈ­ ¼º´É ºñ±³
¿µ¹®Á¦¸ñ(English Title) Performance Comparison of Join Operations Parallelization by using GPGPU
ÀúÀÚ(Author) ÀÌÁ¾¼·   ÀÌ»ó¹é   À̱Ôö   Jong-Sub Lee   Sang-Back Lee   Kyu-Chul Lee  
¿ø¹®¼ö·Ïó(Citation) VOL 34 NO. 03 PP. 0028 ~ 0044 (2018. 12)
Çѱ۳»¿ë
(Korean Abstract)
µ¥ÀÌÅͺ£À̽º ½Ã½ºÅÛ °ü°è ¿¬»êÀÚ Áß¿¡¼­ ¿¬»ê ºñ¿ëÀÌ °¡Àå ºñ½Ñ ¿¬»êÀº Á¶ÀÎ ¿¬»êÀÌ´Ù. ÀϹÝÀûÀ¸·Î CPU ±â¹ÝÀÇ Á¶ÀÎ ¿¬»êÀÇ °æ¿ì¿¡´Â ÇϳªÀÇ Äھ »ç¿ëÇϰųª ¸¹°Ô´Â 16°³ Á¤µµÀÇ Äھ »ç¿ëÇÏ¿© º´·Ä 󸮸¦ Çؼ­ º´·ÄÈ­¿¡ µû¸¥ ¼º´É Çâ»óÀÌ Å©Áö ¾Ê´Ù. ÀÌ¿¡ ¹ÝÇØ, GPGPU(General-Purpose computing on Graphics Processing Units)´Â ¼öõ °³ÀÇ ÇÁ·Î¼¼½Ì À¯´ÖÀ» ÅëÇÑ º´·Ä 󸮰¡ °¡´ÉÇؼ­ Á¶ÀÎ ¿¬»ê ¼öÇà ½Ã°£À» Å©°Ô ´ÜÃàÇÒ ¼ö ÀÖ´Ù. º» ³í¹®¿¡¼­´Â GPGPU ±â¹Ý¿¡¼­ Á¶ÀÎ ¿¬»ê º´·ÄÈ­¸¦ ±¸ÇöÇϱâ À§ÇØ NVIDIAÀÇ CUDA SDK°¡ »ç¿ëµÇ¸ç, CPU ±â¹Ý°ú GPGPU ±â¹Ý¿¡¼­ÀÇ Á¶ÀÎ ¿¬»ê ¼º´ÉÀ» ÃøÁ¤ÇÑ´Ù. »ç¿ëµÇ´Â Á¶ÀÎ ¿¬»êÀº NLJ (Nested Loop Join), SMJ (Merge Join), HJ (Hash Join)À̸ç, GPGPU Àåºñ´Â TITAN Xp, GTX 1080 Ti ¹× GTX1080À» »ç¿ëÇÑ´Ù. CPU ±â¹Ý°ú GPGPU ±â¹ÝÀÇ ¼º´ÉÀ» ºñ±³ÇÏ°í, GPGPU ±â¹ÝÀÇ Á¶ÀÎ ¿¬»ê°ú ÀÌÀü ¿¬±¸ÀÇ ¼º´É°úÀÇ ¼º´ÉÀ» ºñ±³ÇÑ´Ù. ¸¶Áö¸·À¸·Î, ½ÇÇè °á°ú´Â GPGPU ±â¹ÝÀÇ ¼º´ÉÀÌ CPU ±â¹ÝÀÇ ¼º´Éº¸´Ù 6~328 ¹è ºü¸¥ ¼º´ÉÀ» º¸¿´°í ÇâÈÄ ¿¬±¸ÀÇ ¹æÇ⼺¿¡ ´ëÇÏ¿© ÅäÀÇÇÑ´Ù.
¿µ¹®³»¿ë
(English Abstract)
In a database system, the most expensive operation among relational operations is a join operation. Generally, CPU-based join operations uses parallel processing with either 1 core or 16 cores at most, which does not significantly improve the function. On the other hand, GPGPU(General-Purpose computing on Graphics Processing Units) allows parallel processing through thousands of processing units, greatly reducing the time required to perform join operations. Parallelization of the operation using GPGPU uses NVIDIA's CUDA SDK. In this paper, we implement parallelization of the join operation using GPGPU and compare the performances. The used join operations are Nested Loop Join (NLJ), Sort Merge Join (SMJ) and Hash Join (HJ), and GPGPU equipment uses TITAN Xp, GTX 1080 Ti and GTX 1080. We measure and compare the performance of join operations based on CPU and GPGPU. We compare this performance with the performance of the previous study on the join operation based on GPGPU. The results of experiment show that the performance based on GPGPU is 6~328 times faster than the one based on CPU.
Å°¿öµå(Keyword) µ¥ÀÌÅͺ£À̽º ½Ã½ºÅÛ   º´·ÄÈ­ Á¶ÀÎ ¿¬»ê   Database System   GPGPU   CUDA   Parallel Join Operation   Nested-Loop Join   Sort-Merge Join   Hash Join  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå