[B! GPGPU] earth2001yã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

earth2001y id:earth2001y

GPGPUã«é–¢ã™ã‚‹earth2001yã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (21)

${{author_name}}$

{{{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}}

GPUãƒ†ã‚¯ãƒŽãƒã‚¸ãƒ»ã‚«ãƒ³ãƒ•ã‚¡ãƒ¬ãƒ³ã‚¹ | GTC Japan 2012
ã‚¨ãƒŒãƒ“ãƒ‡ã‚£ã‚¢ ã§ã¯ã“ã‚Œã¾ã§ 3å›žã® GPU ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã‚¤ãƒ™ãƒ³ãƒˆã‚’ä¸»å‚¬ã—ã€å¹´ã€…è¦æ¨¡ã‚’æ‹¡å¤§ã—ã¦ãã¾ã—ãŸã€‚4å›žç›®ã¨ãªã‚‹ä»Šå›žã¯å‚åŠ äºˆå®šè€…æ•°1,500åã‚’è¶…ãˆã‚‹æ—¥æœ¬æœ€å¤§ã®HPCé–¢é€£ã‚¤ãƒ™ãƒ³ãƒˆã¨ãªã‚Šã¾ã™ã€‚åŸºèª¿è¬›æ¼”è€…ã«ã¯é•·å¹´ç±³å›½CRAYã§CTOãªã©ã®è¦è·ã‚’å‹™ã‚ãŸã€ç¾ã‚¨ãƒŒãƒ“ãƒ‡ã‚£ã‚¢ç±³å›½æœ¬ç¤¾ Tesla ãƒ“ã‚¸ãƒã‚¹ãƒ¦ãƒ‹ãƒƒãƒˆ CTO ã®ã‚¹ãƒ†ã‚£ãƒ¼ãƒ–ãƒ»ã‚¹ã‚³ãƒƒãƒˆã‚’è¿Žãˆã€GPU ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®æœ€æ–°æƒ…å ±ã€åŠã³æœªæ¥ã®å§¿ã‚’ã”ç´¹ä»‹ã—ã¾ã™ã€‚ ä»Šå¹´ã¯æ–°ã—ã„ GPU ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã§ã‚ã‚‹ Kepler ãŒç™»å ´ã™ã‚‹å¹´ã§ã™ã€‚Kepler ã¯ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯æ€§èƒ½ã€GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°æ€§èƒ½ã‚’å¤§å¹…ã«å¼·åŒ–ã—ãŸã ã‘ã§ã¯ãªãã€ä»®æƒ³åŒ–ãƒ†ã‚¯ãƒŽãƒã‚¸ãƒ¼ã‚’ã‚µãƒãƒ¼ãƒˆã—ãŸå²ä¸Šåˆã®GPUã§ã™ã€‚Keplerã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã‚„ã€GPUä»®æƒ³åŒ–ãƒ†ã‚¯ãƒŽãƒã‚¸ãƒ¼ã§ã‚ã‚‹NVIDIA VGXã¨GeForce Gridã€ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã¨ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒˆã‚’èžåˆã—ãƒ¯ãƒ¼ã‚¯ã‚¹ãƒ†ãƒ¼ã‚·ãƒ§ãƒ³ã‚’
earth2001y 2014/11/26
GTC

nvidia

GPGPU
ãƒªãƒ³ã‚¯
Jetson/Installing CUDA - eLinux.org
earth2001y 2014/11/25
jetson

cuda

GPGPU
ãƒªãƒ³ã‚¯
ã€å¯Œå£«é€šã€‘æ‰‹è»½ã«ãªã£ãŸGPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã€æ§‹é€ è§£æžãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã§åŠ¹æžœã‚’æŽ¢ã‚‹
æ§‹é€ è§£æžãƒ„ãƒ¼ãƒ«ã‚„æµä½“è§£æžãƒ„ãƒ¼ãƒ«ã«ä»£è¡¨ã•ã‚Œã‚‹ã„ã‚ã‚†ã‚‹ã€ŒCAEã€ã®æ´»ç”¨ãŒã€è‡ªå‹•è»Šã‚„ã‚¨ãƒ¬ã‚¯ãƒˆãƒãƒ‹ã‚¯ã‚¹ã‚’ã¯ã˜ã‚ã¨ã™ã‚‹ã•ã¾ã–ã¾ãªåˆ†é‡Žã§é€²ã‚“ã§ã„ã‚‹ä¸ã§ã€NVIDIAç¤¾ã®ã€ŒTesla GPUã‚¢ã‚¯ã‚»ãƒ©ãƒ¬ãƒ¼ã‚¿ã€ã‚’ä½¿ã£ã¦ã‚½ãƒ«ãƒãƒ¼å‡¦ç†ã®æ€§èƒ½å‘ä¸Šã‚’å›³ã‚‹GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã«æ³¨ç›®ãŒé›†ã¾ã£ã¦ã„ã‚‹ã€‚å¯Œå£«é€šã§ã¯æ§‹é€ è§£æžãƒ„ãƒ¼ãƒ«ã®ã€ŒANSYS Mechanicalï¼ˆæœ€æ–°ãƒãƒ¼ã‚¸ãƒ§ãƒ³ 15.0ï¼‰ã€ã‚’å¯¾è±¡ã«GPUã®ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ã‚’å®Ÿæ–½ã—ãŸã€‚çµæžœã®ä¸€éƒ¨ã‚’ç´¹ä»‹ã™ã‚‹ã€‚ CAEã®æ€§èƒ½å‘ä¸Šã®åˆ‡ã‚Šæœã¨ã—ã¦æ•°å¹´å‰ã‹ã‚‰æ³¨ç›®ã•ã‚Œã¦ã„ã‚‹æ‰‹æ³•ã®ã²ã¨ã¤ãŒã€è¶…ä¸¦åˆ—ã‚³ã‚¢ã‚’å†…è”µã—ãŸã‚°ãƒ©ãƒ•ã‚£ã‚¯ã‚¹ãƒ—ãƒã‚»ãƒƒã‚µï¼ˆGPUï¼‰ã‚’ä½¿ã£ã¦ã‚½ãƒ«ãƒãƒ¼ã‚’å®Ÿè¡Œã™ã‚‹ã€ŒGPGPUã€ï¼ˆGeneral-Purpose computing on Graphics Processing Unitsï¼‰ã¾ãŸã¯ã€ŒGPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã€ã§ã‚ã‚‹ã€‚å…·ä½“çš„ã«ã¯NVIDIAç¤¾ã®ã€ŒTesla GPUã‚¢ã‚¯ã‚»ãƒ©ãƒ¬
earth2001y 2014/07/23
GPGPU
ãƒªãƒ³ã‚¯
æ¬¡ä¸–ä»£è‡ªå¾‹ãƒžã‚·ãƒ³ã®ãŸã‚ã® NVIDIA çµ„è¾¼ã¿ã‚·ã‚¹ãƒ†ãƒ
Far Edge Vision AI é–‹ç™ºã‚’å¤‰é©ã™ã‚‹ NVIDIA ã¯ã€å¼·åŠ›ãªãƒžã‚¤ã‚¯ãƒã‚µãƒ¼ãƒ“ã‚¹ã¨ API ã‚’å°Žå…¥ã—ã¦ã€ã“ã‚Œã¾ã§ä»¥ä¸Šã«è¿…é€Ÿãªã‚¨ãƒƒã‚¸ AI ã‚’å®Ÿç¨¼åƒç’°å¢ƒã«ã‚‚ãŸã‚‰ã—ã¾ã™ã€‚ AI ã¯ãƒ“ã‚¸ãƒã‚¹ ãƒ—ãƒã‚»ã‚¹ã®è‡ªå‹•åŒ–ã€ãƒ‡ãƒ¼ã‚¿åˆ†æžã«ã‚ˆã‚‹ã‚¤ãƒ³ã‚µã‚¤ãƒˆã®ç²å¾—ã€é¡§å®¢ã¨å¾“æ¥å“¡ã¨ã®é–¢ã‚ã‚Šãªã©ã€ç«¶äº‰ä¸Šã®å„ªä½æ€§ã‚’æŽ¨é€²ã—ã¦ã„ã¾ã™ã€‚NVIDIAÂ® Jetsonâ„¢ ã¯ã€è‡ªå¾‹ãƒžã‚·ãƒ³ã¨ãã®ä»–ã®çµ„è¾¼ã¿ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ã®ä¸–ç•Œã‚’ãƒªãƒ¼ãƒ‰ã™ã‚‹ãƒ—ãƒ©ãƒƒãƒˆãƒ•ã‚©ãƒ¼ãƒ ã§ã™ã€‚å°åž‹ã§é«˜æ€§èƒ½ãªã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ã® Jetson ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«ã€ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã‚’é«˜é€ŸåŒ–ã™ã‚‹ NVIDIA JetPackâ„¢ SDKã€ã‚»ãƒ³ã‚µãƒ¼ã€SDKã€ã‚µãƒ¼ãƒ“ã‚¹ã€é–‹ç™ºã‚’ã‚¹ãƒ”ãƒ¼ãƒ‰ã‚¢ãƒƒãƒ—ã™ã‚‹è£½å“ã‹ã‚‰ãªã‚‹ã‚¨ã‚³ã‚·ã‚¹ãƒ†ãƒ ãŒå«ã¾ã‚Œã¾ã™ã€‚Jetson ã¯ã€ä»–ã® NVIDIA ãƒ—ãƒ©ãƒƒãƒˆãƒ•ã‚©ãƒ¼ãƒ ã«ä½¿ç”¨ã•ã‚Œã‚‹ã€AI ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã‚„ã‚¯ãƒ©ã‚¦ãƒ‰ãƒã‚¤ãƒ†ã‚£ãƒ– ãƒ¯ãƒ¼ã‚¯ãƒ•ãƒãƒ¼ã¨ã®é–“ã§äº’æ›æ€§ãŒã‚ã‚Šã€ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢
earth2001y 2014/07/18
è²©ç¤¾ã‚µã‚¤ãƒˆã«ã‚ˆã‚Œã°2.4ä¸‡å††ã€‚è‰¯ã„ãªã€‚

GPGPU

nvidia
ãƒªãƒ³ã‚¯
StarPU
A Unified Runtime System for Heterogeneous Multicore Architectures Overview StarPU is a task programming library for hybrid architectures The application provides algorithms and constraints CPU/GPU implementations of tasks A graph of tasks, using either the StarPU's high level GCC plugin pragmas, StarPU's rich C/C++ API, or OpenMP pragmas. StarPU handles run-time concerns Task dependencies Optimiz
earth2001y 2014/01/12
ãƒ˜ãƒ†ãƒã‚¸ãƒ‹ã‚¢ã‚¹è¨ˆç®—æ©Ÿã§ã‚¿ã‚¹ã‚¯ä¸¦åˆ—ã‚’è¡Œã†ãŸã‚ã®ãƒ©ã‚¤ãƒ–ãƒ©ãƒª

HPC

GPGPU
ãƒªãƒ³ã‚¯
20130601.gpgpuå‹‰å¼·ä¼š ãƒãƒ¼ã‚¿ãƒ–ãƒ«ã‚¹ãƒ‘ã‚³ãƒ³è£½ä½œé¡›æœ«è¨˜
2. 2013/06/01 (c) 2013 kalab1998 2 è‡ªå·±ç´¹ä»‹ â— ã‹ã‚‰ã¼ã€€ã“ã¨ã€€æ¸¥ç¾Žæ¸…éš† â— twitter: @kalab1998 Kiyotaka Atsumi labã¯1998å¹´ã‹ã‚‰å§‹ã¾ã£ãŸï¼Ž â— facebook: https://www.facebook.com/kalab1998 â— éˆ´é¹¿é«˜å°‚æƒ…å ±å‡¦ç†ã‚»ãƒ³ã‚¿ãƒ¼å°‚ä»»ã®å‡†æ•™æŽˆ ã©ã¡ã‚‰ã‹ã¨è¨€ãˆã°æƒ…å ±ã‚»ã‚ãƒ¥ãƒªãƒ†ã‚£å±‹ã•ã‚“
earth2001y 2013/06/07
ãƒãƒƒãƒ†ãƒªãƒ¼ã®é‹ç”¨ã‚„ã°ã„

ã“ã‚Œã¯ã™ã”ã„

GPGPU
ãƒªãƒ³ã‚¯
CUDA vs. Phi: Phi Programming for CUDA Developers
Currently we allow the following HTML tags in comments: Single tags These tags can be used alone and don't need an ending tag. <br> Defines a single line break <hr> Defines a horizontal line Matching tags These require an ending tag - e.g. <i>italic text</i> <a> Defines an anchor <b> Defines bold text <big> Defines big text <blockquote> Defines a long quotation <caption> Defines a table caption <c
earth2001y 2012/12/20
cuda

GPGPU

intel

xeonphi

HPC
ãƒªãƒ³ã‚¯
GPGPU Advent Calendarã‚„ã‚ŠãŸã„ã§ã™ - ãƒ‘ãƒ©ãƒ¬ãƒ«ã«æ‹ã—ã¦
æ¶ˆãˆãŸè¨˜äº‹ã‚’ã©ã†ã«ã‹æ›¸ãç›´ãã†ã¨é ‘å¼µã‚‹ GK110ã‚’æè¼‰ã—ãŸTeslaã“ã¨Tesla K20ãŒå‡ºã¦ã¾ã„ã‚Šã¾ã—ãŸ ä¸–ç•Œæœ€å¤§ãƒ¢ãƒ³ã‚¹ã‚¿ãƒ¼ãƒãƒƒãƒ—ãŒä¸Šä½ã‚’å¸å·»ã™ã‚‹ã®ã‹ï¼ï¼Ÿ IntelãŒæ”¾ã¤Xeon phiã¨ã„ã†åˆºå®¢ãŒGPGPUã«ã¨ã©ã‚ã‚’åˆºã™ã®ã‹ï¼ï¼Ÿã¶ã£ã¡ã‚ƒã‘x86ã®Cellã˜ã‚ƒããƒ¼ã‹ ãã‚“ãªGPUæ—¥å’Œãª2012ã€ç·æ‹¬ã¨ã—ã¦ã€ã‚³ãƒ³ãƒ‹ãƒãƒ¯GPUãªæ„Ÿã˜ã§ã€æ§˜ã€…ã«å¿œç”¨ã•ã‚Œã¦ã„ã‚‹GPGPUã‚’æŽ¢ã—ãŸã‚Šè‡ªåˆ†ã§ä½œã£ã¦ã¿ãŸã‚Šã™ã‚‹ã‚ˆã†ãªæ„Ÿã˜ã®Advent Calendarã‚’é–‹å‚¬ã—ãŸã„ã¨æ€ã£ãŸã‚Šã—ã¾ã™ ãƒã‚¿ã¨ã—ã¦æŽ¨å¥¨ã™ã‚‹ã®ã¯ã€ GPGPUã®è«–æ–‡(æ—¥è‹±å•ã‚ãš) GPGPUé–‹ç™ºç’°å¢ƒã®ç´¹ä»‹(LLã¸ã®Wrapperãªã©) GPGPUã§ã“ã‚“ãªã®ã‚„ã£ã¦ã¿ãŸ ã¨ã„ã†ã‚ˆã†ãªã‚‚ã®ã§ã™ã€‚ GPGPUã£ã¦é›£ã—ã„ã‚“ã˜ã‚ƒãªã„ã®ï¼Ÿ ã¨ã„ã†ã‚ãªãŸã‚‚ã€ã“ã‚Œã‚’æ©Ÿã«èªã‚“ã ã‚Šæ›¸ã„ãŸã‚Šã—ã¦ã¿ã‚‹ã¨ã€CPUã§ã¯å‘³ã‚ãˆãªã‹ã£ãŸæ–°ã—ã„å¿«æ„Ÿã«å‡ºä¼šãˆã‚‹ã‹ã‚‚ã—ã‚Œmâ€¦ æ€ªç‰©ã¨
earth2001y 2012/11/22
é›†ã¾ã£ãŸã‚‰é¢ç™½ãã†ã€é›†ã¾ã‚‹ã®ã‹ï¼Ÿ

GPGPU

adventcalendar
ãƒªãƒ³ã‚¯
ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã®ã‚³ãƒ³ãƒ‘ã‚¤ãƒ©ã€ŒLLVMã€ãŒã€NVIDIA GPUã‚’ã‚µãƒãƒ¼ãƒˆ
NVIDIAã¯ã€ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã®ã‚³ãƒ³ãƒ‘ã‚¤ãƒ©ã€ŒLLVMã€ãŒã€NVIDIA GPUã‚’ã‚µãƒãƒ¼ãƒˆã—ãŸã“ã¨ã‚’ã€9æ—¥ï¼ˆç¾åœ°æ™‚é–“ï¼‰ã«ç™ºè¡¨ã—ãŸã€‚ã€ŒLLVMã€ã«ã‚ˆã‚‹NVIDIA GPUã®ã‚µãƒãƒ¼ãƒˆã«ã‚ˆã£ã¦ã€GPUã‚¢ã‚¯ã‚»ãƒ©ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ã‚’åˆ©ç”¨å¯èƒ½ãªãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°è¨€èªžãŒå¤§å¹…ã«å¢—ãˆã‚‹ã¨ã¨ã‚‚ã«ã€ãã®ãƒ¡ãƒªãƒƒãƒˆã‚’äº«å—ã§ãã‚‹ç ”ç©¶è€…ã‚„ç‹¬ç«‹ç³»ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ãƒ»ãƒ™ãƒ³ãƒ€ãƒ¼ï¼ˆISVï¼‰ãŒã€åŠ‡çš„ã«å¢—ãˆã‚‹ã“ã¨ã«ãªã‚‹ã€‚ ã€ŒLLVMã€ã¯ã€å¹…åºƒãä½¿ã‚ã‚Œã¦ã„ã‚‹ã‚ªãƒ¼ãƒ—ãƒ³ã‚½ãƒ¼ã‚¹ã®ã‚³ãƒ³ãƒ‘ã‚¤ãƒ©ãƒ»ã‚¤ãƒ³ãƒ•ãƒ©ã‚¹ãƒˆãƒ©ã‚¯ãƒãƒ£ã§ã€ãƒ¢ã‚¸ãƒ¥ãƒ¼ãƒ«åž‹ã®è¨è¨ˆã«ã‚ˆã£ã¦ã‚µãƒãƒ¼ãƒˆã™ã‚‹ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°è¨€èªžã‚„ãƒ—ãƒã‚»ãƒƒã‚µãƒ»ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã‚’ç°¡å˜ã«è¿½åŠ ã§ãã‚‹ã€‚C/C++ã€Objective-Cã€Fortranã€Adaã€Haskellã€Javaãƒã‚¤ãƒˆã‚³ãƒ¼ãƒ‰ã€ Pythonã€Rubyã€ActionScriptã€GLSLã€Rustãªã©ã€å¤šå½©ãªãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°è¨€èªžã¨ãƒ•ãƒãƒ³ãƒˆã‚¨ãƒ³ãƒ‰ã«å¯¾å¿œã—ã¦ã„ã‚‹ã€‚ åŒç¤¾ã¯ã€ã€ŒLL
earth2001y 2012/05/14
ã†ãŠã€ãƒžã‚¸ã‹ï¼

GPGPU

ã‚³ãƒ³ãƒ‘ã‚¤ãƒ©
ãƒªãƒ³ã‚¯
ã€äºˆå‘Šã€‘é–¢æ±GPGPUå‹‰å¼·ä¼šé–‹å‚¬ã—(ãŸã„ã¨æ€ã£ã¦ã„)ã¾ã™ï¼ - ãƒ‘ãƒ©ãƒ¬ãƒ«ã«æ‹ã—ã¦
ã™ã¹ã¦ã¯ã“ã®ç™ºè¨€ãŒãã£ã‹ã‘ã§ã‚ã‚‹ã¾ã™ã€‚ ã•ã‚‰ã«ã„ã†ã¨ã€ã“ã®ä»¶ã¯ã€@_likrã•ã‚“ã® ãŒãã£ã‹ã‘ã§ã™ã€‚ @_likrã•ã‚“ã¨ã¯ä»¥å‰ã«æ±äº¬ã®æŸæ‰€ã§ãŠä¼šã„ã—ãŸã“ã¨ãŒã‚ã‚Š(ã¨ã„ã†æ›¸ãæ–¹ã ã¨è‹¥å¹²èª¤è§£ã‚’æ‹›ãæ„Ÿã˜ï¼Ÿ) ã€ãªã‚“ã‹ååˆºã¾ã§ã‚‚ã‚‰ã£ã¡ã‚ƒã£ãŸæ°—ã‚‚ã—ã¦ã€å®Ÿéš›ã®ã¨ã“ã‚ã¯é¢è˜ã‚ã£ãŸã‚Šã™ã‚‹ã¨ã‹ãã†ã„ã†é–‘è©±ä¼‘é¡Œã€‚ ã•ã¦ãŠãã€‚ é–¢è¥¿ã§ã‚„ã‚‹ãªã‚‰é–¢æ±ã§ã‚‚ã‚„ã‚‰ãªãã‚ƒï¼ã¨ã„ã†ã“ã¨ã§ é–¢æ±GPGPUå‹‰å¼·ä¼šé–‹å‚¬ã—ãŸã„ã¨æ€ã£ã¦ã„ã¾ã™ ã¨ã„ã†ã‹ä¸ç”¨æ„ãªç™ºè¨€ã—ãŸã‚‰å¤–å €ãŒåŸ‹ã¾ã£ã¦ã„ãŸã§ã”ã–ã‚‹ ã¤ãã¾ã—ã¦ã¯ç™ºè¡¨è€…ã®æ–¹ã‚’å‹Ÿé›†ã—ãŸã„ã®ã§æˆ‘ã“ãã¯çš„ãªæ–¹ã¯æ˜¯éžåä¹—ã‚Šã‚’ä¸Šã’ã¦ãã ã•ã„ãŠé¡˜ã„ã—ã¾ã™ ã‚¿ãƒ¼ã‚²ãƒƒãƒˆã¨ã—ã¦ã©ã“ã‚’ç›®æŒ‡ã™ã®ã‹ã¯ã€å®Ÿéš›ç™ºè¡¨è€…ã®æ–¹ã€…ãŒã©ã‚Œãã‚‰ã„ã®æ¿ƒã„è©±ã‚’ã—ã¦ãã ã•ã‚‹ã‹ã«ã‹ã‹ã£ã¦ã„ã¾ã™ï¼ ç¾åœ¨è€ƒãˆã¦ã„ã‚‹å‹‰å¼·ä¼šã®ãƒ‡ã‚£ãƒ†ã‚£ãƒ¼ãƒ«ã§ã™ãŒ æ±äº¬ã®ã©ã“ã‹(è²¸ã—ä¼šè°å®¤ãªã©ã‚’æŽ¢ã™) ãƒ†ãƒ¼ãƒžã¯GPGPUã«ã¾ã¤ã‚ã‚‹ã“ã¨ å‚åŠ äººæ•°ã¯å ´æ‰€ã«å¿œã˜ã¦ æ™‚é–“
earth2001y 2012/04/12
èˆˆå‘³ã‚¢ãƒªã‚¢ãƒªã€‚ãœã²ã¨ã‚‚é€±æœ«ã«é–‹å‚¬ã—ã¦ã„ãŸã ããŸãã€‚ã€‚ã€‚

GPGPU

event
ãƒªãƒ³ã‚¯
About NVIDIA
ç‘æ³¢å¤§å¦ã€ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ãƒ¼ã«GPUã‚’æè¼‰ã—ã€å®‡å®™ç‰©ç†å¦ã‚„ãƒ©ã‚¤ãƒ•ã‚µã‚¤ã‚¨ãƒ³ã‚¹ã¨ã„ã£ãŸæ§˜ã€…ãªé›£ã—ã„åˆ†é‡Žã«ãŠã‘ã‚‹ç ”ç©¶ã‚’æŽ¨é€² 1,000å°è¶…ã®NVIDIA Tesla GPUã§CPUã‚µãƒ¼ãƒã‚’é«˜é€ŸåŒ–ã—ã€ è¶…ä¸¦åˆ—ã‚¯ãƒ©ã‚¹ã‚¿ã®ã‚·ãƒ³ã‚°ãƒ«ãƒŽãƒ¼ãƒ‰ãƒ»ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ã§ãƒˆãƒƒãƒ—ã‚¯ãƒ©ã‚¹ã®æ€§èƒ½ã‚’é”æˆ 2012å¹´3æœˆ22æ—¥ - NVIDIA(æœ¬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€Nasdaqï¼šNVDA)ã¯æœ¬æ—¥ã€å›½ç«‹å¤§å¦æ³•äººç‘æ³¢å¤§å¦ãŒå¯†çµåˆä¸¦åˆ—æ¼”ç®—åŠ é€Ÿæ©Ÿæ§‹å®Ÿè¨¼ã‚·ã‚¹ãƒ†ãƒ ã€ŒHA-PACSã€ã«1000å°ä»¥ä¸Šã‚‚ã®NVIDIA GPUã‚’æè¼‰ã—ã€ã•ã¾ã–ã¾ãªåˆ†é‡Žã«ãŠã‘ã‚‹ç ”ç©¶ã‚’é«˜é€ŸåŒ–ã™ã‚‹ã“ã¨ã«ãªã£ãŸã¨ç™ºè¡¨ã—ã¾ã—ãŸã€‚ ç‘æ³¢å¤§å¦è¨ˆç®—ç§‘å¦ç ”ç©¶ã‚»ãƒ³ã‚¿ãƒ¼ã«è¨ç½®ã•ã‚Œã¦ã„ã‚‹HA-PACSã‚·ã‚¹ãƒ†ãƒ ã¯1,072å°ã®NVIDIAÂ® Teslaâ„¢ M2090 GPUã§ã‚¢ãƒƒãƒ—ã‚°ãƒ¬ãƒ¼ãƒ‰ã•ã‚Œã€ç†è«–
earth2001y 2012/03/28
HPC

GPGPU

nvidia
ãƒªãƒ³ã‚¯
ã€ãƒ¬ãƒãƒ¼ãƒˆã€‘ã“ã‚Œã‹ã‚‰å…ˆã®ç§‘å¦ã‚’é¢ç™½ã„ä¸–ç•Œã«å¤‰ãˆã¦ã„ã - GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°2010 | ã‚¨ãƒ³ã‚¿ãƒ¼ãƒ—ãƒ©ã‚¤ã‚º | ãƒžã‚¤ã‚³ãƒŸã‚¸ãƒ£ãƒ¼ãƒŠãƒ«
earth2001y 2010/07/20
ã€Œãƒ¬ã‚¬ã‚·ãƒ¼ã‚³ãƒ¼ãƒ‰ã®é«˜é€ŸåŒ–ã€ã€Œå¤ã„ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã‹ã‚‰ä¸¦åˆ—ã«æœ€é©åŒ–ã•ã‚ŒãŸæ–°ã—ã„ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢ã¸ã®æ›¸ãæ›ãˆã®ä¿ƒé€²ã€ã€Œã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚„æ¼”ç®—æ–¹æ³•ã®å†è€ƒã«ã‚ˆã‚‹ã€å¾“æ¥ã¨ã¯ç•°ãªã‚‹è¦³ç‚¹ã«ã¨ã‚‹æ–°ãŸãªã‚¢ãƒ—ãƒãƒ¼ãƒæ‰‹æ³•ã®ç¢ºç«‹ã€

GPGPU

hpc
ãƒªãƒ³ã‚¯
ã€å¾Œè—¤å¼˜èŒ‚ã®Weeklyæµ·å¤–ãƒ‹ãƒ¥ãƒ¼ã‚¹ã€‘ GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã¨ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚¹ã«åˆ†åŒ–ã™ã‚‹NVIDIA GPU
earth2001y 2010/07/07
GPGPU
ãƒªãƒ³ã‚¯
æ±å·¥å¤§ã€GPUã‚¹ãƒ‘ã‚³ãƒ³ã‚’æ´»ç”¨ã—æ¬¡ä¸–ä»£æ°—è±¡äºˆæ¸¬ãƒ¢ãƒ‡ãƒ«ã®è¨ˆç®—é€Ÿåº¦å‘ä¸Šã‚’é”æˆ | ã‚¨ãƒ³ã‚¿ãƒ¼ãƒ—ãƒ©ã‚¤ã‚º | ãƒžã‚¤ã‚³ãƒŸã‚¸ãƒ£ãƒ¼ãƒŠãƒ«
æ±äº¬å·¥æ¥å¤§å¦ã¯3æœˆ24æ—¥ã€æ°—è±¡åºãŒé–‹ç™ºã‚’é€²ã‚ã¦ã„ã‚‹æ¬¡æœŸæ°—è±¡äºˆå ±å‘ã‘æ¬¡ä¸–ä»£æ°—è±¡ãƒ¢ãƒ‡ãƒ«ã€ŒASUCAã€ã‚’åŒå¤§ã®ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã€ŒTSUBAME 1.2ã€ã«æè¼‰ã—ãŸGPUã‚’æ´»ç”¨ã—ã¦ã€ãã®ã™ã¹ã¦ã‚’GPUã§å‡¦ç†ã™ã‚‹ã“ã¨ã§å¾“æ¥æ¼”ç®—æ–¹å¼ã«æ¯”ã¹ã¦æ¼”ç®—é€Ÿåº¦ã®é«˜é€ŸåŒ–ã‚’å®Ÿç¾ã—ãŸã“ã¨ã‚’æ˜Žã‚‰ã‹ã«ã—ãŸã€‚ æ—¥æœ¬ã§GPUã‚¹ãƒ‘ã‚³ãƒ³ã®é“ç‹ã‚’åˆ‡ã‚Šé–‹ã„ã¦ããŸTSUBAME1.2 ä¸€èˆ¬çš„ãªæ°—è±¡ãƒ¢ãƒ‡ãƒ«ã¯å¤§æ°—åœ§ã¨é‡åŠ›ã®ã¤ã‚Šã‚ã„ã®å…¼ãåˆã„ã‹ã‚‰ä¸Šä¸‹ã«ã¯é¢¨ãŒå¹ã‹ãªã„ã“ã¨ãŒå‰æã¨ãªã£ã¦ã„ã‚‹ãŒã€äºˆæ¸¬ç¯„å›²ã®é¢ç©ã‚’ãƒ¡ãƒƒã‚·ãƒ¥çŠ¶ã«åŒºåˆ†ã‘ã—ã€ãã‚Œã‚’ç´°ã‹ãã™ã‚Œã°ã™ã‚‹ã»ã©ã€ä¾‹ãˆã°å°é¢¨ãªã©ã‚’ä¾‹ã«å–ã‚‹ã¨ã€å¹³è¡¡ã«åŠ›ã¯åƒã‹ãªããªã‚Šã€é›²ã®ä¸Šä¸‹é‹å‹•ãªã©ã‚‚å«ã‚ãŸè¨ˆç®—ãŒå¿…è¦ã¨ãªã£ã¦ãã‚‹3æ¬¡å…ƒéžé™åŠ›å¦å¹³è¡¡ãƒ¢ãƒ‡ãƒ«ã¨ãªã‚‹ã€‚ã¾ãŸã€ãã†ã—ãŸé›²ã®è§£åƒã‚’æ°—è±¡äºˆå ±ã«å¿œç”¨ã™ã‚‹ãŸã‚ã«ã¯å°é¢¨ã®ã‚ˆã†ãª2000kmã«ã‚‚ãŠã‚ˆã¶ç¯„å›²ã®ã‚‚ã®ã‹ã‚‰ã€2-3kmç¨‹åº¦ã®ç¯„å›²ã§åŽã¾ã‚‹ç«œå·»ã‚„é›†ä¸è±ªé›¨ã¨ã„ã£ãŸã‚‚ã®ã¾ã§
earth2001y 2010/05/26
ã‚¹ãƒ‘ã‚³ãƒ³

GPGPU

HPC
ãƒªãƒ³ã‚¯
æ±å·¥å¤§ã€æ°—è±¡åºã®æ¬¡ä¸–ä»£æ°—è±¡ãƒ¢ãƒ‡ãƒ«ã®ãƒ•ãƒ«GPUåŒ–ã«æˆåŠŸ
earth2001y 2010/03/24
GPGPU

ã‚·ãƒŸãƒ¥ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³
ãƒªãƒ³ã‚¯
å¯Œå£«é€šé•·é‡Žã‚·ã‚¹ãƒ†ãƒ ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ãƒªãƒ³ã‚°ï¼Œã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚¹ã®ãƒãƒ¼ãƒ‰ã§ä¼ç†±è§£æžã®è¨ˆç®—ã‚’é«˜é€ŸåŒ–
å¯Œå£«é€šé•·é‡Žã‚·ã‚¹ãƒ†ãƒ ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ãƒªãƒ³ã‚°ï¼ˆæœ¬ç¤¾é•·é‡Žå¸‚ï¼‰ã¯ï¼ŒåŒç¤¾ã®ãƒªãƒ•ãƒãƒ¼ç‚‰ä¼ç†±è§£æžã‚·ã‚¹ãƒ†ãƒ ã€ŒReflowPlusã€ã®è¨ˆç®—å‡¦ç†ã‚’3ï½ž10å€ã«é«˜é€ŸåŒ–ã—ãŸæ–°ç‰ˆã€ŒReflowPlus V02L03ã€ã‚’2010å¹´2æœˆ1æ—¥ã‹ã‚‰è²©å£²ã™ã‚‹ã€‚ReflowPlusã¯2009å¹´6æœˆã«ç™ºå£²ã—ãŸãƒ—ãƒªãƒ³ãƒˆå›žè·¯åŸºæ¿ï¼ˆPCBï¼‰ã®ãƒªãƒ•ãƒãƒ¼å·¥ç¨‹ã®ã‚·ãƒŸãƒ¥ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ã«ç”¨ã„ã‚‹ã‚½ãƒ•ãƒˆã€‚è¨ˆç®—å‡¦ç†ã‚’ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚¹å‡¦ç†ç”¨ãƒ—ãƒã‚»ãƒƒã‚µï¼ˆGraphics Processing Unitï¼ŒGPUï¼‰ã«å®Ÿè¡Œã•ã›ã‚‹ã“ã¨ã§ï¼Œå¤§å¹…ã«é«˜é€ŸåŒ–ã§ããŸã€‚
earth2001y 2010/01/25
GPGPU
ãƒªãƒ³ã‚¯
GPGPU ãŒã‚¢ãƒ„ã„ã€‚ä¸€ç•ªã§ã¯ãªãä¸€ç•ªä¹—ã‚Šã«ãªã‚ã†ã€‚ - æ¦è”µé‡Žæ—¥è¨˜
PFI ã‚»ãƒŸãƒŠãƒ¼ GPU ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®ç¾çŠ¶ã¨ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®æœªæ¥ã‚’è¦‹ã‚‹ã€‚äºˆæƒ³ä»¥ä¸Šã«é¥ã‹ã«ãŠã‚‚ã—ã‚ã‹ã£ãŸã€‚èª¬æ˜Žã‚‚åˆ†ã‹ã‚Šã‚„ã™ã„ã—ãªã«ã‚ˆã‚Š @nushio ã•ã‚“ã®è©±ã—æ–¹ãŒå¿ƒåº•æ¥½ã—ãã†ãªã®ãŒã¨ã¦ã‚‚ã‚ˆã„ :-) ã‚¨ãƒ³ã‚¸ãƒ‹ã‚¢ã«ã›ã‚ˆç ”ç©¶è€…ã«ã›ã‚ˆã€ä¸€ç•ªå¤§äº‹ãªã®ã¯è‡ªåˆ†ã®ã‚„ã£ã¦ã„ã‚‹ã“ã¨ã«æƒ…ç†±ã‚’ã‹ã‘ã‚‰ã‚Œã‚‹ã‹ã€æ¥½ã—ã‚“ã§ã„ã‚‹ã‹ã€ã¨ã„ã†ã“ã¨ã«å°½ãã‚‹ã¨æ€ã†ã€‚ä¸Šè¨˜ã®ãƒªãƒ³ã‚¯ã¯æ”¾æ˜ ã•ã‚ŒãŸã‚‚ã®ã®éŒ²ç”»ãªã®ã ãŒã€ã‚¹ãƒ©ã‚¤ãƒ‰ã¯ã“ã¡ã‚‰ã€‚ãƒˆãƒ¼ã‚¯ã‚‚ã¡ã‚‡ã†ã©1æ™‚é–“ç¨‹åº¦ãªã®ã§ã€æ™‚é–“ãŒã‚ã‚Œã°ã‚¹ãƒ©ã‚¤ãƒ‰ã ã‘ã§ã¯ãªãã€ãƒˆãƒ¼ã‚¯ã®æœ€åˆã®æ•°åˆ†ã ã‘ã§ã‚‚è¦‹ã‚‹(ã¨ã„ã†ã‹èžã)ã“ã¨ã‚’ãŠå‹§ã‚ã™ã‚‹ã€‚å…ˆæ—¥ã®ã‚¦ã‚§ãƒ–å¦ä¼šã®ã¨ãã‚‚æ€ã£ãŸãŒã€æ±äº¬ã«ã„ãªãã¦ã‚‚ã“ã†ã„ã†ã®ãŒèžã‘ã‚‹æ™‚ä»£ã«ãªã£ã¦ã€ã™ã”ãä¾¿åˆ©ã ãªã‚ã€‚(ã‚ã¨ã‹ã‚‰è¦‹ã‚‹ã“ã¨ã‚‚ã§ãã‚‹ã—ã€æ–‡å—é€šã‚Šæ™‚ç©ºã‚’è¶…ãˆã¦ã„ã‚‹) GPGPU ã£ã¦ãªã‚“ãªã®ã‚ˆã€ã¨ã„ã†ã¨ã€ã“ã‚Œã¯ General Purpose computing
earth2001y 2009/12/14
ã‚¹ãƒ‘ã‚³ãƒ³

GPGPU
ãƒªãƒ³ã‚¯
GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®ç¾çŠ¶ã¨ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®æœªæ¥
PFI Seminar 20091210
earth2001y 2009/12/10
ustream

gpgpu
ãƒªãƒ³ã‚¯
2009/12/10 GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®ç¾çŠ¶ã¨ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®æœªæ¥
PFIç¤¾å†…ã‚»ãƒŸãƒŠãƒ¼ 2009å¹´12æœˆ10æ—¥ 20:00-21:00(äºˆå®š) GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®ç¾çŠ¶ã¨ã‚¹ãƒ¼ãƒ‘ãƒ¼ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®æœªæ¥ ç™ºè¡¨è€…: æ‘ä¸» å´‡è¡Œï¼ˆãƒ—ãƒªãƒ•ã‚¡ãƒ¼ãƒ‰ã‚¤ãƒ³ãƒ•ãƒ©ã‚¹ãƒˆãƒ©ã‚¯ãƒãƒ£ãƒ¼ ç ”ç©¶é–‹ç™ºéƒ¨é–€ãƒ»äº¬éƒ½å¤§å¦å¤§å¦é™¢ã€€ç‰©ç†å¦ç¬¬äºŒæ•™å®¤ï¼‰ ã‚»ãƒŸãƒŠãƒ¼éŒ²ç”»URL: http://www.ustream.tv/recorded/2837689 ã“ã®ã‚¹ãƒ©ã‚¤ãƒ‰ã¯ã€ç™ºè¡¨å¾Œã«ã¿ãªã•ã¾ã‹ã‚‰ã„ãŸã ã„ãŸè²´é‡ãªæ„è¦‹ã‚’ã‚‚ã¨ã«æ”¹è¨‚ã—ãŸç‰ˆã§ã™ã€‚ç™ºè¡¨æ™‚ç‚¹ã§ã®ç‰ˆã¯ã“ã¡ã‚‰: http://www.slideshare.net/pfi/20091210-gpu-2735685Read less
earth2001y 2009/12/10
GPGPU

ã“ã‚Œã¯ã™ã”ã„
ãƒªãƒ³ã‚¯
http://twitter.com/yomuraya/status/1058334012
earth2001y 2008/12/15
GPGPUç‰ˆã®SETIã¨ãªï¼Ÿ

GPGPU
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸