Ç¥ÁØÈ­ Âü¿©¾È³»

TTAÀÇ Ç¥ÁØÇöȲ

Ȩ > Ç¥ÁØÈ­ °³¿ä > TTAÀÇ Ç¥ÁØÇöȲ

Ç¥ÁعøÈ£ TTAK.KO-11.0291 ±¸Ç¥ÁعøÈ£
Á¦°³Á¤ÀÏ 2021-12-08 ÃÑÆäÀÌÁö 21
ÇѱÛÇ¥ÁØ¸í ´ÙÁß NPU¿¡¼­ ´ÙÁß ½Å°æ¸Á Ãß·ÐÀ» À§ÇÑ ¿î¿µÃ¼Á¦ ¿ä±¸»çÇ×
¿µ¹®Ç¥Áظí Operating System Requirements for Inferring Multiple Neural Nets on Multiple NPUs
Çѱ۳»¿ë¿ä¾à ÀÌ Ç¥ÁØÀº ½Å°æ¸ÁÀ» Ãß·ÐÇϱâ À§ÇÑ ÀÓº£µðµå ¿î¿µÃ¼Á¦ÀÇ ¿ä±¸»çÇ×À» ±â¼úÇÑ´Ù. ÇØ´ç ¿î¿µÃ¼Á¦´Â ¸¹Àº ±â´ÉÀ» Á¦°øÇÒ ¼ö ÀÖÁö¸¸ ÀÌ Ç¥ÁØ¿¡¼­´Â ½Å°æ¸ÁÀ» Ãß·ÐÇϱâ À§ÇÑ ½Å°æ¸Á ¸ðµ¨ÀÇ ·Îµù/¾ð·Îµù ±â´É, ½Å°æ¸Á Ä¿³Î ·Îµù/¾ð·Îµù ±â´É, ½Å°æ¸Á Ãß·ÐÀ» À§ÇÑ Å½ºÅ© ±â¹ÝÀÇ Ãß·Ð ±â´É, ½Å°æ¸Á ŽºÅ© ½ºÄÉÁÙ¸µ ±â´É, È¿À²ÀûÀÎ ½Å°æ¸Á ŽºÅ© ½ºÄÉÁÙ¸µÀ» À§ÇÑ ¸ð´ÏÅ͸µ ±â´É¿¡ ´ëÇÏ¿© ³íÀÇÇÑ´Ù.
¿µ¹®³»¿ë¿ä¾à This standard describes the requirements of embedded operating systems for inferring multiple neural nets. The operating systems for neural net inference have various functions, but the reference model of this standard first discusses the neuralnet model, kernel loading/unloading function, inference function using neuralnet task, neuralnet task scheduling function, and system monitoring function.
±¹Á¦Ç¥ÁØ
°ü·ÃÆÄÀÏ TTAK_[2].KO-11.0291.pdf TTAK_[2].KO-11.0291.pdf            

ÀÌÀü
À¥ ¿ÀÇ API º¸¾È ¿ä±¸ »çÇ×
´ÙÀ½
ÇÇ½Ì »ç°í¿¡ ´ëÇÑ º¸¾È Áöħ