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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document : 3 / 5 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) ÅëÇո޸𸮸¦ ÀÌ¿ëÇÑ ÀÓº£µðµå ȯ°æ¿¡¼­ÀÇ µö·¯´× ÇÁ·¹ÀÓ¿öÅ© ¼º´É °³¼±°ú Æò°¡
¿µ¹®Á¦¸ñ(English Title) Performance Enhancement and Evaluation of a Deep Learning Framework on Embedded Systems using Unified Memory
ÀúÀÚ(Author) À̹ÎÇР  °­¿ìö   Minhak Lee   Woochul Kang  
¿ø¹®¼ö·Ïó(Citation) VOL 23 NO. 07 PP. 0417 ~ 0423 (2017. 07)
Çѱ۳»¿ë
(Korean Abstract)
ÃÖ±Ù, µö·¯´×À» »ç¿ë °¡´ÉÇÑ ÀÓº£µðµå µð¹ÙÀ̽º°¡ »ó¿ëÈ­µÊ¿¡ µû¶ó ÀÓº£µðµå ½Ã½ºÅÛ ¿µ¿ª¿¡¼­µµ µö·¯´× È°¿ë¿¡ ´ëÇÑ ´Ù¾çÇÑ ¿¬±¸°¡ ÁøÇàµÇ°í ÀÖ´Ù. ±×·¯³ª ÀÓº£µðµå ½Ã½ºÅÛÀ» °í¼º´É PC ȯ°æ°ú ºñ±³ÇÏ¸é »ó´ëÀûÀ¸·Î Àú»ç¾çÀÇ CPU/GPU ÇÁ·Î¼¼¼­¿Í ¸Þ¸ð¸®¸¦ žÀçÇÏ°í ÀÖÀ¸¹Ç·Î µö·¯´× ±â¼úÀÇ Àû¿ë¿¡ À־ ¸¹Àº Á¦¾àÀÌ ÀÖ´Ù. º» ³í¹®¿¡¼­´Â ´Ù¾çÇÑ ÃֽŠµö·¯´× ³×Æ®¿öÅ©µéÀ» ÀÓº£µðµå µð¹ÙÀ̽º¿¡ Àû¿ëÇßÀ»¶§ÀÇ ¼º´ÉÀ» ½Ã°£°ú Àü·ÂÀ̶ó´Â °üÁ¡¿¡¼­ ½ÇÇèÀûÀ¸·Î Æò°¡ÇÑ´Ù. ¶ÇÇÑ, È£½ºÆ® CPU¿Í GPU µð¹ÙÀ̽º°£ÀÇ ¸Þ¸ð¸®¸¦ °øÀ¯ÇÏ´Â ÀÓº£µðµå ½Ã½ºÅÛµéÀÇ ¾ÆÅ°ÅØóÀûÀΠƯ¼ºÀ» ÀÌ¿ëÇÏ¿© ¸Þ¸ð¸® º¹»ç¸¦ ÁÙÀÓÀ¸·Î½á ½Ç½Ã°£ ¼º´É°ú ÀúÀü·Â¼ºÀ» ³ôÀÌ´Â ¹æ¹ýÀ» Á¦½ÃÇÑ´Ù. Á¦¾ÈµÈ ¹æ¹ýÀº ´ëÇ¥ÀûÀÎ °ø°³ µö·¯´× ÇÁ·¹ÀÓ¿öÅ©ÀÎ Caffe¸¦ ¼öÁ¤ÇÏ¿© ±¸ÇöµÇ¾úÀ¸¸ç, ÀÓº£µðµå GPU¸¦ žÀçÇÑ NVIDIA Jetson TK1¿¡¼­ ¼º´ÉÆò°¡ µÇ¾ú´Ù. ½ÇÇè°á°ú, ´ëºÎºÐÀÇ µö·¯´× ³×Æ®¿öÅ©¿¡¼­ ¶Ñ·ÇÇÑ ¼º´ÉÇâ»óÀ» °üÂûÇÒ ¼ö ÀÖ¾ú´Ù. ƯÈ÷, ¸Þ¸ð¸® »ç¿ë·®ÀÌ ³ôÀº AlexNet ¿¡¼­ ¾à 33%ÀÇ À̹ÌÁö ÀÎ½Ä ¼Óµµ ´ÜÃà°ú 50%ÀÇ ¼Òºñ Àü·Â·® °¨¼Ò¸¦ °üÂûÇÒ ¼ö ÀÖ¾ú´Ù.
¿µ¹®³»¿ë
(English Abstract)
Recently, many embedded devices that have the computing capability required for deep learning have become available; hence, many new applications using these devices are emerging. However, these embedded devices have an architecture different from that of PCs and highperformance servers. In this paper, we propose a method that improves the performance of deep-learning framework by considering the architecture of an embedded device that shares memory between the CPU and the GPU. The proposed method is implemented in Caffe, an open-source deep-learning framework, and is evaluated on an NVIDIA Jetson TK1 embedded device. In the experiment, we investigate the image recognition performance of several state-of-the-art deeplearning networks, including AlexNet, VGGNet, and GoogLeNet. Our results show that the proposed method can achieve significant performance gain. For instance, in AlexNet, we could reduce image recognition latency by about 33% and energy consumption by about 50%.
Å°¿öµå(Keyword) ÀÓº£µðµå ½Ã½ºÅÛ   GPU   µö·¯´×   ÀúÀü·Â   À̹ÌÁö ½Äº°   embedded system   GPU   deep learning   low power   image recognition  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå