[B! GPGPU] nikkoroã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

nikkoro id:nikkoro

GPGPUã«é–¢ã™ã‚‹nikkoroã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (11)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

ãƒˆãƒ¼ã‚¿ãƒ«ãƒ»ãƒ‡ã‚£ã‚¹ã‚¯ãƒãƒ¼ã‚¸ãƒ£ãƒ»ã‚µã‚¤ãƒˆï¼ˆäº‹å®Ÿã‚’ã‚ã‚Šã®ã¾ã¾ã«ï¼‰
ã“ã®Wikiã¯ã€é–‹ç™ºã®ã€Œç”Ÿã€ãƒ‡ãƒ¼ã‚¿ã®æƒ…å ±é–‹ç¤ºã‚’ç›®çš„ã¨ã—ã¦ã„ã¾ã™ã€‚ å¾“ã£ã¦ä»•æ§˜ãªã©ã®æƒ…å ±ã¯é »ç¹ã«æ›´æ–°ã•ã‚Œã¾ã™ã®ã§ã€å‡ºè·ã•ã‚Œã‚‹è£½å“ã®ä»•æ§˜ã¨ã¯ç•°ãªã‚‹å ´åˆã®æ–¹ãŒå¤šã„ã§ã™ã€‚æ›´æ–°ã•ã‚ŒãŸéŽåŽ»ã®æƒ…å ±ã¯ å±¥æ´ ã‹ã‚‰å‚ç…§ã™ã‚‹ã“ã¨ãŒå¯èƒ½ã§ã™ã€‚ ã€Œãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã€ã‚»ã‚¯ã‚·ãƒ§ãƒ³ã¯ã€ãŠã‚‚ã«å½“ç¤¾ã®è£½å“ã§ä½¿ç”¨ã•ã‚Œã‚‹äº‹ãŒå¤šã„ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚„ã€æ–°ãŸã«é©ç”¨ã•ã‚Œã‚‹OSã«ã¤ã„ã¦ã®æƒ…å ±ã‚’è¨˜è¼‰ã—ã¾ã™ã€‚ ã€ŒæŠ€è¡“æƒ…å ±ã€ã‚»ã‚¯ã‚·ãƒ§ãƒ³ã¯ã€HPCã‚·ã‚¹ãƒ†ãƒ ã‚ºç§˜è”µã®æŠ€è¡“æƒ…å ±ã‚’å…¬é–‹ã—ã¦ã„ã¾ã™ã€‚ ã“ã®ã‚µã‚¤ãƒˆã¸ã®ã”æ„è¦‹ã¯ã‚µã‚¤ãƒˆã¸ã®æ›¸ãè¾¼ã¿ã®ã¿ã§ãŠå—ã‘ã„ãŸã—ã¾ã™ã€‚ã¾ãŸã€è¿”ä¿¡ã®æ–¹æ³•ã¯ã‚µã‚¤ãƒˆã¸ã®æ›¸ãè¾¼ã¿ã®ã¿ã¨ã•ã›ã¦ã„ãŸã ãã¾ã™ã€‚ ã“ã®ã‚µã‚¤ãƒˆã®æƒ…å ±ã¯ã€ç‰¹è¨˜ã®ãªã„ã‹ãŽã‚Šã€å„ãƒšãƒ¼ã‚¸ã®åŸ·ç†æ™‚ç‚¹ã«ãŠã‘ã‚‹çŠ¶æ³ã«åŸºã¥ã„ã¦è¨˜è¿°ã•ã‚Œã¦ã„ã¾ã™ã€‚åŸ·ç†å¾Œã«çŠ¶æ³ç‰ãŒå¤‰åŒ–ã—ã€é–²è¦§æ™‚ã®æœ€æ–°ã®çŠ¶æ³ã¨ã¯å†…å®¹ãŒç•°ãªã‚‹å¯èƒ½æ€§ãŒã‚ã‚‹ã“ã¨ã‚’ã€ã‚ã‚‰ã‹ã˜ã‚ã”æ‰¿çŸ¥ãŠããã ã•ã„ã€‚ æ³¨ç›®æƒ…å ±ï¼ æœ¬æ—¥ã®å‡ºè· 2009
nikkoro 2012/10/16
GPGPU

programming
ãƒªãƒ³ã‚¯
WebCL
WebCL extension for Firefox, providing direct GPU access from JavaScriptThe WebCL project exposes OpenCL into JavaScript, allowing web developers to tap into the massive parallel computing resources of modern GPUs and multicore CPUs. This, when combined with WebGL and other emerging standards, enables entirely new categories of interactive web apps, such as photo editing, video processing, visuali
nikkoro 2011/12/23
gpgpu

webcl
ãƒªãƒ³ã‚¯
ã‚¢ãƒžã‚¾ãƒ³ã€ã€ŒEC2ã€ã§GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã‚’æä¾›é–‹å§‹
ç±³ã‚¢ãƒžã‚¾ãƒ³ãƒ»ãƒ‰ãƒƒãƒˆãƒ»ã‚³ãƒ ã®Webã‚µãƒ¼ãƒ“ã‚¹éƒ¨é–€ã‚¢ãƒžã‚¾ãƒ³ãƒ»ã‚¦ã‚§ãƒ–ãƒ»ã‚µãƒ¼ãƒ“ã‚·ã‚ºã¯2010å¹´11æœˆ15æ—¥ã€GPUï¼ˆã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯å‡¦ç†ãƒ—ãƒã‚»ãƒƒã‚µï¼‰ã‚’ä½¿ã£ãŸIaaSï¼ˆã‚¤ãƒ³ãƒ•ãƒ©ã‚¹ãƒˆãƒ©ã‚¯ãƒãƒ£ãƒ»ã‚¢ã‚ºãƒ»ã‚¢ãƒ»ã‚µãƒ¼ãƒ“ã‚¹ï¼‰ã‚’æä¾›é–‹å§‹ã—ãŸã€‚ç±³ã‚¨ãƒŒãƒ“ãƒ‡ã‚£ã‚¢ã®GPUã§ã‚ã‚‹ã€ŒTesla M2050ã€ã‚’ä½¿ã£ãŸè¨ˆç®—æ©Ÿå‡¦ç†ã‚µãƒ¼ãƒ“ã‚¹ã‚’ã€ã‚¢ãƒžã‚¾ãƒ³ã®ã€ŒEC2ã€çµŒç”±ã§æä¾›ã™ã‚‹ã€‚GPUã‚’ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚¹å‡¦ç†ä»¥å¤–ã«æ´»ç”¨ã™ã‚‹ã€ŒGPGPUï¼ˆæ±Žç”¨GPUï¼‰ã€ã¨å‘¼ã¶åˆ†é‡ŽãŒã€ã‚¯ãƒ©ã‚¦ãƒ‰ã‚µãƒ¼ãƒ“ã‚¹ã«ã‚‚ç™»å ´ã™ã‚‹ã€‚ æ–°ãƒ¡ãƒ‹ãƒ¥ãƒ¼ã¯ã€Œã‚¯ãƒ©ã‚¹ã‚¿ãƒ¼GPUã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹ã€ã€‚ä¸€ã¤ã®ã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹ï¼ˆåˆ©ç”¨å˜ä½ï¼‰å½“ãŸã‚Šã€Tesla M2050ã‚’2å€‹ã€ã‚¤ãƒ³ãƒ†ãƒ«ã®Xeon X5570ã‚’2å€‹ä½¿ã„ã€ãƒ¡ãƒ¢ãƒªãƒ¼22Gãƒã‚¤ãƒˆã€ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸1690Gãƒã‚¤ãƒˆã€10Gãƒ“ãƒƒãƒˆ/ç§’ã®ã‚¤ãƒ¼ã‚µãƒãƒƒãƒˆã‚’æä¾›ã™ã‚‹ã€‚1ã‚¤ãƒ³ã‚¹ã‚¿ãƒ³ã‚¹ã®åˆ©ç”¨æ–™é‡‘ã¯1æ™‚é–“å½“ãŸã‚Š2ãƒ‰ãƒ«10ã‚»ãƒ³ãƒˆã€‚ GPGPUã¨ã¯CPUã¨GPUã‚’çµ„ã¿åˆã‚ã›ã€æµ®å‹•å°æ•°
nikkoro 2010/11/16
cloud computing

GPGPU
ãƒªãƒ³ã‚¯
æ—¥æœ¬IBMã€GPGPUã‚’æ´»ç”¨ã—ãŸHPCã‚¯ãƒ©ã‚¦ãƒ‰ã‚µãƒ¼ãƒ“ã‚¹ã‚’æä¾›é–‹å§‹
å°åˆ·ã™ã‚‹ ãƒ¡ãƒ¼ãƒ«ã§é€ã‚‹ ãƒ†ã‚ã‚¹ãƒˆ HTML é›»åæ›¸ç± PDF ãƒ€ã‚¦ãƒ³ãƒãƒ¼ãƒ‰ ãƒ†ã‚ã‚¹ãƒˆ é›»åæ›¸ç± PDF ã‚¯ãƒªãƒƒãƒ—ã—ãŸè¨˜äº‹ã‚’Myãƒšãƒ¼ã‚¸ã‹ã‚‰èªã‚€ã“ã¨ãŒã§ãã¾ã™ æ—¥æœ¬IBMã¯11æœˆ2æ—¥ã€ã‚¯ãƒ©ã‚¦ãƒ‰ã‚µãƒ¼ãƒ“ã‚¹ã€ŒIBM Computing on Demandã€ã‚’æ‹¡å……ã—ã€å¤§é‡ãƒ‡ãƒ¼ã‚¿ã®ä¸¦åˆ—å‡¦ç†ã«å„ªã‚ŒãŸãƒã‚¤ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ï¼ˆHPCï¼‰ç’°å¢ƒã‚’æä¾›ã™ã‚‹ã‚µãƒ¼ãƒ“ã‚¹ã‚’åŒæ—¥ã‚ˆã‚Šé–‹å§‹ã™ã‚‹ã¨ç™ºè¡¨ã—ãŸã€‚ IBM Computing on Demandã¯ã€IBMã®ãƒ‡ãƒ¼ã‚¿ã‚»ãƒ³ã‚¿ãƒ¼ã«è¨ç½®ã•ã‚Œã€HPCå‘ã‘ã«æ§‹æˆã•ã‚ŒãŸã‚µãƒ¼ãƒãŠã‚ˆã³ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸ã‚’ã‚»ã‚ãƒ¥ãƒªãƒ†ã‚£ã®ç¢ºä¿ã•ã‚ŒãŸãƒãƒƒãƒˆãƒ¯ãƒ¼ã‚¯ã‚’ä»‹ã—ã¦æä¾›ã™ã‚‹åŒç¤¾ã®ã‚¯ãƒ©ã‚¦ãƒ‰ã‚µãƒ¼ãƒ“ã‚¹ã€‚ãƒ¦ãƒ¼ã‚¶ãƒ¼ã®ç™ºæ³¨ã‹ã‚‰1é€±é–“ä»¥å†…ã«å¿…è¦ãªè³‡æºã‚’æä¾›ã™ã‚‹ã¨ã„ã†ã€‚ãƒ¦ãƒ¼ã‚¶ãƒ¼ã¯å¿…è¦ãªã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã‚’ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã—ã€ã‚¤ãƒ³ãƒˆãƒ©ãƒãƒƒãƒˆä¸Šã®è³‡æºã¨ã—ã¦ç®¡ç†ã§ãã‚‹ã€‚ç¹å¿™æœŸã¸ã®å¯¾å¿œã‚„çŸæœŸé–“ã§ã®æ–°è£½å“é–‹ç™ºãªã©ã§ä¸€æ™‚çš„ãªãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã®
nikkoro 2010/11/03
GPGPU

IBM

cloud computing
ãƒªãƒ³ã‚¯
ã€ã‚¤ãƒ™ãƒ³ãƒˆãƒ¬ãƒãƒ¼ãƒˆã€‘ ã€GTC 2010ãƒ¬ãƒãƒ¼ãƒˆã€‘ã€åŸºèª¿è¬›æ¼”ç·¨ã€‘ ã‚¹ã‚¿ãƒ³ãƒ•ã‚©ãƒ¼ãƒ‰å¤§æ•™æŽˆå…¼GoogleæŠ€è¡“è€…ãŒç„¡äººè‡ªå¾‹èµ°è¡Œè‡ªå‹•è»Šå®Ÿç¾ã¸ã®é“ã‚’èªžã‚‹
nikkoro 2010/10/26
GPGPU

Google
ãƒªãƒ³ã‚¯
Developer
Welcome to the GPGPU.org Developer Resources page. Here you will find a wealth of information on the GPGPU programming, including tutorials, utility code, examples, useful links and much more. All of the code here is open source. Jump to GPGPU Programming Conference Tutorials Recommended Reading Material GPGPU Programming The GPGPU programming landscape has rapidly evolved over the past several ye
nikkoro 2010/10/26
GPGPU
ãƒªãƒ³ã‚¯
ã€PC Watchã€‘ GPU Technology Conference 2010Â ãƒ¬ãƒãƒ¼ãƒˆãƒªãƒ³ã‚¯é›†
ã€9/24ã€‘ ã€GTC 2010ãƒ¬ãƒãƒ¼ãƒˆã€‘ ã€æŠ€è¡“ã‚»ãƒƒã‚·ãƒ§ãƒ³ç·¨ã€‘GPUãƒ™ãƒ¼ã‚¹ã®é«˜é€Ÿå†™çœŸç·¨é›†ã‚½ãƒ•ãƒˆã€ŒMusemageã€ ã€œGPUã‚’ä½¿ã£ãŸRAIDæŠ€è¡“ãªã©ã‚‚ ç¬ åŽŸä¸€è¼ã®ãƒ¦ãƒ“ã‚ã‚¿ã‚¹æƒ…å ±å±€ CUDAã®ä¸–ç•Œã‚’åºƒã’ã‚‹â€œCUDA Everywhereâ€æ§‹æƒ³ ã€œãƒ¢ãƒ“ãƒªãƒ†ã‚£ãƒ‡ãƒã‚¤ã‚¹ã§ã¯x86ã§ã¯ãªãARMã«è³ã‘ã‚‹NVIDIA ã€9/22ã€‘ ã€GTC 2010ãƒ¬ãƒãƒ¼ãƒˆã€‘ ã€ã‚¸ã‚§ãƒ³ãƒ»ã‚¹ãƒ³ãƒ»ãƒ•ã‚¢ãƒ³æ°åŸºèª¿è¬›æ¼”ã€‘2011å¹´ã«ã€ŒKeplerã€2013å¹´ã«ã¯ã€ŒMaxwellã€ã¨ã„ã†GPUã‚’æŠ•å…¥ ã€œMaxwellã¯Teslaã®16å€ã®é›»åŠ›æ€§èƒ½ã‚’å®Ÿç¾ ã€9/21ã€‘ ã€GTC 2010ãƒ¬ãƒãƒ¼ãƒˆã€‘ NVIDIAä¸»å‚¬ã®GPU æŠ€è¡“ä¼šè°ã€ŒGTCã€ãŒé–‹å¹•
nikkoro 2010/10/22
GPGPU
ãƒªãƒ³ã‚¯
å¾Œè—¤å¼˜èŒ‚ã®Weeklyæµ·å¤–ãƒ‹ãƒ¥ãƒ¼ã‚¹
â—äººæã‚’å¼•ãå…¥ã‚Œã¦å¤‰èº«ã‚’ç¹°ã‚Šè¿”ã™NVIDIA NVIDIAã®æŠ€è¡“çš„ãªæ–¹å‘æ€§ã¯ã€åŒç¤¾ã«ã©ã‚“ãªäººç‰©ãŒåŠ ã‚ã£ãŸã‹ã«ã‚ˆã£ã¦ã‚ã‹ã‚‹ã€‚NVIDIA GPUã®æœ€å¤§ã®è»¢æ©Ÿã¨ãªã£ãŸã€ŒGeForce 8800(G80)ã€ã®èƒŒå¾Œã«ã¯ã€ä¸¦åˆ—ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã®ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒˆã§ã‚ã‚‹John Nickolls(ã‚¸ãƒ§ãƒ³ãƒ»ãƒ‹ã‚³ãƒ«ã‚º)æ°(ç¾Director of Architecture, Nvidia)ãŒã„ãŸã€‚Nickollsæ°ã¯ã€è¶…ä¸¦åˆ—SIMD(Single Instruction, Multiple Data)ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã€ŒMasPar Computerã€ã®ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒˆã ã£ãŸäººç‰©ã€‚MasParã®ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã‚’ã€éƒ¨åˆ†çš„ã«GPUã«æŒã¡è¾¼ã‚“ã ã€‚ã€Œ(MasParã¨G80ã«ã¯)é•ã†éƒ¨åˆ†ã‚‚è‰²ã€…ã‚ã‚‹ãŒã€å…ƒã«ãªã£ã¦ã„ã‚‹ã‚¢ã‚¤ãƒ‡ã‚¢ã¯éžå¸¸ã«ä¼¼ã¦ã„ã‚‹ã€ã¨Nickollsè‡ªèº«ã‚‚èªžã£ã¦ã„ãŸã€‚ NVIDIAã®äººæé›†ã‚ã¯GPUãƒãƒ¼ãƒ‰çµ¡ã¿ã ã‘ã§ã¯ãª
nikkoro 2010/10/22
GPGPU
ãƒªãƒ³ã‚¯
100å€ã®ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚‚å¤¢ã§ã¯ãªã„ã€‚æ¬¡ä¸–ä»£ã‚¹ãƒ‘ã‚³ãƒ³ã‚’ä½“ç¾ã™ã‚‹GPU+CPUãƒã‚¤ãƒ–ãƒªãƒƒãƒ‰ã‚½ãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã€‚
100å€ã®ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã‚‚å¤¢ã§ã¯ãªã„ã€‚ æ¬¡ä¸–ä»£ã‚¹ãƒ‘ã‚³ãƒ³ã‚’ä½“ç¾ã™ã‚‹GPU+CPUãƒã‚¤ãƒ–ãƒªãƒƒãƒ‰ã‚½ãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã€‚ æ—¥æœ¬ã®å¦è¡“æ©Ÿé–¢ã«ã‚ˆã‚‹ä¸–ç•Œæœ€é€Ÿã‚’ç›®æŒ‡ã—ãŸã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ã¸ã®å–ã‚Šçµ„ã¿ãŒå¤§ããªæœŸå¾…ã‚’é›†ã‚ã¦ã„ã‚‹ãŒï¼ŒGPUï¼ˆGraphics Processing Unitï¼‰ã‚’æœ¬æ ¼çš„ã«æŽ¡ç”¨ã—ãŸäº‹ä¾‹ã¨ã—ã¦ã‚‚æ³¨ç›®ã«å€¤ã™ã‚‹ã‚‚ã®ã ã€‚ä»Šå¾Œï¼Œä¼æ¥ã®HPCï¼ˆHigh Performance Computingï¼‰é ˜åŸŸã§ã‚‚GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ãŒé€²å±•ã™ã‚‹ã“ã¨ã¯é–“é•ã„ãªã„ã ã‚ã†ã€‚2010å¹´10æœˆã«æ—¥æœ¬HPãŒç™ºè¡¨ã—ãŸã€ŒHP ProLiant SL6500ã‚·ãƒªãƒ¼ã‚ºã€ã¯ï¼ŒHPCã«ç‰¹åŒ–ã—ãŸæ–°ã—ã„ã‚µãƒ¼ãƒãƒ¼ãƒ©ã‚¤ãƒ³ã§ã‚ã‚‹ã¨ã¨ã‚‚ã«ï¼ŒGPU+CPUåŒæ–¹ã®è‰¯ã•ã‚’å¼•ãå‡ºã—ãŸãƒã‚¤ãƒ–ãƒªãƒƒãƒ‰è£½å“ã ã€‚ä¼æ¥ã§ã®GPUåˆ©ç”¨ã®ç¾çŠ¶ã¨ï¼Œè£½å“ã®ç‰¹é•·ã«ã¤ã„ã¦æ—¥æœ¬HPã®ã‚ãƒ¼ãƒ‘ãƒ¼ã‚½ãƒ³2äººã«è©³ã—ãèžã“ã†ã€‚ ã‚µã‚¤ãƒåŒ–ã‹ã‚‰çµ±åˆåŒ–ã¸ï¼Œãã—ã¦æ¬¡ãªã‚‹ä¸€æ‰‹ã¯ï¼Ÿ æ˜¨ä»Šï¼Œè£½é€ ãƒ»é‡‘
nikkoro 2010/10/14
GPGPU
ãƒªãƒ³ã‚¯
About NVIDIA
CUDAã‚’ã‚µãƒãƒ¼ãƒˆã—ãŸOpenCVã€æ¬¡ä¸–ä»£ãƒãƒœãƒ†ã‚£ã‚¯ã‚¹ã‹ã‚‰å®‰å…¨ãªè‡ªå‹•è»Šã¾ã§ã€ ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ãƒ“ã‚¸ãƒ§ãƒ³ã‚’ãƒ™ãƒ¼ã‚¹ã¨ã—ãŸãƒ¡ã‚¤ãƒ³ã‚¹ãƒˆãƒªãƒ¼ãƒ ç”¨ ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã®å®Ÿç”¨åŒ–ã«é“ã‚’æ‹“ã 2010å¹´9æœˆ23æ—¥ - ã‚«ãƒªãƒ•ã‚©ãƒ«ãƒ‹ã‚¢å·žã‚µãƒ³ã‚¿ã‚¯ãƒ©ãƒ© - NVIDIAï¼ˆæœ¬ç¤¾ï¼šç±³å›½ã‚«ãƒªãƒ•ã‚©ãƒ«ãƒ‹ã‚¢å·žã‚µãƒ³ã‚¿ã‚¯ãƒ©ãƒ©ã€ç¤¾é•·å…¼CEOï¼š ã‚¸ã‚§ãƒ³ã‚¹ãƒ³ãƒ»ãƒ•ã‚¢ãƒ³(Jen-Hsun Huang)ã€Nasdaqï¼šNVDAï¼‰ã¯æœ¬æ—¥ã€ãƒãƒœãƒ†ã‚£ã‚¯ã‚¹ã€è‡ªå‹•è»Šã€åŒ»ç™‚ã€æ¶ˆè²»è€…è£½å“ã€ã‚»ã‚ãƒ¥ãƒªãƒ†ã‚£ã€è£½é€ ã€ç ”ç©¶ãªã©ã®åˆ†é‡Žã§é«˜åº¦ãªã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã‚’é–‹ç™ºã™ã‚‹éš›ã«äººæ°—ã®é«˜ã„ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ãƒ“ã‚¸ãƒ§ãƒ³ãƒ»ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã€OpenCVãŒCUDAã‚’ã‚µãƒãƒ¼ãƒˆã™ã‚‹ã¨ç™ºè¡¨ã—ã¾ã—ãŸã€‚ OpenCVã§GPUã‚¢ã‚¯ã‚»ãƒ©ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ãŒåˆ©ç”¨ã§ãã‚‹ã‚ˆã†ã«ãªã‚‹ã¨ã€ä»Šå¾Œã€é–‹ç™ºæ™‚ã«ã€ç²¾åº¦ãŒé«˜ãé«˜åº¦ãªOpenCVã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’é«˜è§£åƒåº¦ã®ç”»åƒã«å¯¾ã—ã¦ãƒªã‚¢ãƒ«ã‚¿ã‚¤ãƒ ã«å®Ÿè¡Œã™ã‚‹ã“ã¨ãŒå¯èƒ½ã«ãªã‚Šã¾ã™ã€‚ã—ã‹ã‚‚ã€æ¶ˆè²»é›»åŠ›
nikkoro 2010/09/28
OpenCV

CUDA

GPGPU
ãƒªãƒ³ã‚¯
ã€ã‚¤ãƒ™ãƒ³ãƒˆãƒ¬ãƒãƒ¼ãƒˆã€‘ ã€GTC 2010ãƒ¬ãƒãƒ¼ãƒˆã€‘ ã€æŠ€è¡“ã‚»ãƒƒã‚·ãƒ§ãƒ³ç·¨ã€‘GPUãƒ™ãƒ¼ã‚¹ã®é«˜é€Ÿå†™çœŸç·¨é›†ã‚½ãƒ•ãƒˆã€ŒMusemageã€ ï½žGPUã‚’ä½¿ã£ãŸRAIDæŠ€è¡“ãªã©ã‚‚
nikkoro 2010/09/24
GPGPU

gpu
ãƒªãƒ³ã‚¯
1