[B! GPU] gigs123ã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

gigs123 id:gigs123

GPUã«é–¢ã™ã‚‹gigs123ã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (17)

${{author_name}}$

{{{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}}

GPUï¼ˆã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ãƒœãƒ¼ãƒ‰ï¼‰æ€§èƒ½æ¯”è¼ƒè¡¨ã€2026å¹´æœ€æ–°ç‰ˆã€‘
æŽ²è¼‰ã®ä¾¡æ ¼ã¯ã€æ›´æ–°æ™‚ç‚¹ã§ã®ãŠãŠã‚ˆãã®å¸‚å ´æœ€å®‰å€¤ä¾¡æ ¼ã¨ãªã£ã¦ã„ã¾ã™ã€‚ç¾åœ¨ã¯ç•°ãªã‚‹å¯èƒ½æ€§ãŒã‚ã‚‹ãŸã‚æ³¨æ„ã—ã¦ãã ã•ã„ã€‚ã¾ãŸã€åŒã˜GPUåã§ã‚‚è£½å“ã«ã‚ˆã£ã¦ä»•æ§˜ãŒç•°ãªã‚‹å ´åˆãŒã‚ã‚‹ãŸã‚æ³¨æ„ã—ã¦ãã ã•ã„ï¼ˆç‰¹ã«ãƒ¡ãƒ¢ãƒªã‚„ã‚¯ãƒãƒƒã‚¯ï¼‰ã€‚GPUåã®ãƒªãƒ³ã‚¯ã¯Amazonã®ã‚‚ã®ã§ã™ã€‚ã€Œã‚°ãƒ©ãƒœã€ã¯ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ãƒœãƒ¼ãƒ‰ã®ç•¥ç§°ã§ã™ã€‚ 3DMark Time Spyï¼ˆDX12 / 1440p ã®ã‚²ãƒ¼ãƒ æ€§èƒ½ï¼‰ ç¾çŠ¶Windowsã®PC ã‚²ãƒ¼ãƒ ã§ä¸»æµã¨ã„ãˆã‚‹DirectX 12ï¼ˆDX12ï¼‰ã®ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ãƒ†ã‚¹ãƒˆã§ã™ã€‚è§£åƒåº¦ã€ŒWQHDï¼ˆ2560Ã—1440ï¼‰ã€ã®Graphicsã‚¹ã‚³ã‚¢ã«ãªã‚Šã¾ã™ã€‚3DMark Steel Nomadï¼ˆDX12 / 4K ã®ã‚²ãƒ¼ãƒ æ€§èƒ½ï¼‰ ç¾çŠ¶Windowsã®PC ã‚²ãƒ¼ãƒ ã§ä¸»æµã¨ã„ãˆã‚‹DirectX 12ï¼ˆDX12ï¼‰ã®ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯ãƒ†ã‚¹ãƒˆã§ã™ã€‚è§£åƒåº¦ã€Œ4Kï¼ˆ3840Ã—2160ï¼‰ã€ã®Graphicsã‚¹ã‚³ã‚¢ã«ãªã‚Šã¾ã™ã€‚
gigs123 2025/08/12
GPU

æ€§èƒ½

æ¯”è¼ƒ
ãƒªãƒ³ã‚¯
Stability AIã€å°‚ç”¨GPUä¸è¦ã®ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°å‘ã‘ç”ŸæˆAIã€ŒStable Code 3Bã€ã‚’å…¬é–‹ï¼ãƒ©ãƒƒãƒ—ãƒˆãƒƒãƒ—PCã§ã‚‚ã‚ªãƒ•ãƒ©ã‚¤ãƒ³ã§ã®ãƒªã‚¢ãƒ«ã‚¿ã‚¤ãƒ å®Ÿè¡ŒãŒå¯èƒ½
gigs123 2024/01/17
GPU

PC

AI
ãƒªãƒ³ã‚¯
OpenInterpreter ã‚’Metalã‚’ä½¿ã£ãŸãƒãƒ¼ã‚«ãƒ«ã®Code Llamaã§ä½¿ã£ã¦ã¿ãŸï½œalexweberk
OpenInterpreter ã¯ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã ã¨ GPT-4 ãŒä½¿ã‚ã‚Œã‚‹ãŒã€ãƒãƒ¼ã‚«ãƒ«ã® Code Llama ã‚’ä½¿ã†ã“ã¨ã‚‚ã§ãã‚‹ã¨ã„ã†ã“ã¨ã§ã€ è©¦ã—ã«è¨å®šã—ã¦ä½¿ã£ã¦ã¿ã¾ã—ãŸã€‚ è¨å®šã‚’ã™ã‚‹ä¸Šã§ä½•ç‚¹ã‹ã¤ã¾ã¥ã„ãŸã®ã§ã€è§£æ±ºã«ç¹‹ãŒã£ãŸã‚‚ã®ã‚’ãƒ¡ãƒ¢ã—ã¾ã™ã€‚ ä»Šå›žä½¿ã£ãŸãƒãƒ¼ãƒ‰ã‚¦ã‚§ã‚¢ç’°å¢ƒã¯ã€M1 Macbook Pro 16GB ã§ã™ã€‚ ãƒãƒ¼ã‚«ãƒ«ã® Code Llama ã‚’ä½¿ã†ã«ã¯OpenInterpreter ã¯ã€ãƒãƒ¼ã‚«ãƒ«ã® Code Llama ã‚’ä½¿ã†ã“ã¨ãŒã§ãã¾ã™ã€‚ ãƒãƒ¼ã‚«ãƒ«ã® Code Llama ã‚’ä½¿ã†ã«ã¯ã€ä»¥ä¸‹ã®ã‚ˆã†ã«ã—ã¾ã™ã€‚ interpreter --localä½¿ã„ãŸã„ãƒ¢ãƒ‡ãƒ«ã®ãƒ‘ãƒ©ãƒ¼ãƒ¡ãƒ¼ã‚¿ã‚’èžã‹ã‚Œã‚‹ã®ã§ã€ä»Šå›žã¯ã€Œ7Bã€ã€ŒMediumã€ã€ŒGPU ã‚ã‚Šã€ã‚’é¸æŠžã—ã¾ã—ãŸã€‚ Open Interpreter will use Code Llama for local execution. Use
gigs123 2023/09/28
python

æ©Ÿæ¢°å¦ç¿’

GPU
ãƒªãƒ³ã‚¯
Accelerating Standard C++ with GPUs Using stdpar | NVIDIA Technical Blog
NVC++ is the first compiler to support GPU-accelerated Standard C++ with no language extensions, pragmas, directives, or non-standard libraries, allowing for porta ble code that can be accelerated on NVIDIA GPUs.C++17 introduced higher-level parallelism features, including execution policies like std::execution::par and std::execution::par_unseq, which enable parallelization of Standard Library alg
gigs123 2020/08/09
GPU

CPU
ãƒªãƒ³ã‚¯
145mmã¨ã‚³ãƒ³ãƒ‘ã‚¯ãƒˆãªGeForce GTX 1050ãŒELSAã‹ã‚‰
sponsored ã‚¢ã‚¹ã‚ãƒ¼ç·¨é›†éƒ¨ãŒBacklogå°Žå…¥ã§å¾—ãŸæˆæžœã¯ï¼Ÿ è§£æ±ºã§ããªã‹ã£ãŸæ‚©ã¿ã¯ï¼Ÿ å…¨éƒ¨è©±ã—ã¾ã™ å¿–åº¦ãªã—ã§èªžã‚ã†ã€€Backlogãƒ¦ãƒ¼ã‚¶ãƒ¼æ´1å¹´åŠã€ã€Œãƒãƒ¼ãƒ ã§åƒãã€ã£ã¦ã‚„ã£ã±ã‚Šé›£ã—ã„ï¼ sponsored ãƒªãƒ“ãƒ³ã‚°ã®ã‚¤ãƒ³ãƒ†ãƒªã‚¢ã¨ã—ã¦ã‚‚é•å’Œæ„Ÿãªã—ï¼ å£²ã‚Œç‹ãƒŸãƒ‰ãƒ«ã‚¿ãƒ¯ãƒ¼PCã‚±ãƒ¼ã‚¹ã®å‰é¢ãŒæœ¨è£½ã«ï¼ çµ„ã¿ã‚„ã™ã„å·¥å¤«å¤šæ•°ã®ã€ŒFRAME 4000Dã€ãŒã‚ˆã‚Šã‚ªã‚·ãƒ£ãƒ¬ã«ãªã£ãŸã‚ˆ sponsored STYLE-15FHA21-R7A-UCSXã‚’ãƒ¬ãƒ“ãƒ¥ãƒ¼ 165Hzæ¶²æ™¶&Ryzen AI 7 350æè¼‰Copilot+ PCã®æ€§èƒ½ã‚’æ¤œè¨¼ã€è»½ã‚ã®ã‚²ãƒ¼ãƒ ãªã‚‰1920Ã—1200ãƒ‰ãƒƒãƒˆã§å¹³å‡100fpsè¶… sponsored æ ¼å®‰ãƒžã‚¶ãƒ¼ã§ãƒã‚¤ã‚¨ãƒ³ãƒ‰CPUã¯ç„¡ç†ï¼Ÿ ASRockã®B860ãƒžã‚¶ãƒ¼ã§Core Ultra 7 270K Plusã‚’å‹•ã‹ã—ã¦ã¿ãŸ sponsored JN-i245G280F-HSP
gigs123 2016/11/19
GPU
ãƒªãƒ³ã‚¯
ã€ŒGeForce GTX 1050/1050Tiã€ã¯ä½Žäºˆç®—ã‚²ãƒ¼ãƒžãƒ¼ã®è²¡å¸ƒã‚’ã“ã˜é–‹ã‘ã‚‹ã‹ï¼Ÿ (1/6)
10æœˆ18æ—¥ã€NVIDIAã¯Pascalä¸–ä»£ã®ã‚¨ãƒ³ãƒˆãƒªãƒ¼ãƒŸãƒ‰ãƒ«GPUã€ŒGeForce GTX 1050ã€ï¼ˆä»¥ä¸‹ã€GTX 1050ã€ãŠã‚ˆã³ãã®ä¸Šä½ãƒ¢ãƒ‡ãƒ«ã€ŒGeForce GTX 1050Tiã€ï¼ˆä»¥ä¸‹ã€GTX 1050Tiï¼‰ã‚’ç™ºè¡¨ã€‚10æœˆ25æ—¥22æ™‚ã‚ˆã‚ŠGTX 1050Tiã®è²©å£²ãŒé–‹å§‹ã•ã‚ŒãŸã€‚ ã¡ãªã¿ã«GTX 1050ã¯å¾®å¦™ã«å‡ºè¶³ãŒé…ãã€ç™ºå£²ã¯25æ—¥ä»¥é™ã€11æœˆ8æ—¥ã¾ã§ã«é †æ¬¡ç™ºå£²ã•ã‚Œã‚‹ã¨ã®ã“ã¨ã ã€‚ GTX 1050/1050Tiç™»å ´ã®èƒŒæ™¯ã¨ã—ã¦ã€ã“ã“æ•°å¹´ã§PCå‘ã‘ã‚²ãƒ¼ãƒ ã®é–‹ç™ºæ•°ã®ä¼¸ã³ã¨ã€PCç”¨ã‚²ãƒ¼ãƒ ã®é‡é‡åŒ–ãŒåŠ é€Ÿã—ã¤ã¤ã‚ã‚‹ç‚¹ã‚’NVIDIAãŒã‚ã’ã¦ã„ã‚‹ã€‚ å…ˆä»£ã®ãƒŸãƒ‰ãƒ«ãƒ¬ãƒ³ã‚¸GPUã€ŒGeForce GTX 950ã€ã¯ã€eã‚¹ãƒãƒ¼ãƒ„ã‚²ãƒ¼ãƒžãƒ¼ã®è²·ã„æ›¿ãˆã‚’å–šèµ·ã—ã‚ˆã†ã¨ã—ãŸè£½å“ã ã£ãŸãŒã€ä»Šå›žã®GTX 1050/1050Tiã¯ãã‚Œã‚’PCã‚²ãƒ¼ãƒžãƒ¼å…¨ä½“ã«æ‹¡å¤§ã—ã‚ˆã†ã¨ã„ã†è£½å“ãªã®ã ã€‚ ã‚¹ãƒšãƒƒã‚¯ã¯å¾Œã§è§£èª¬ã™ã‚‹ãŒã€ã©ã¡
gigs123 2016/11/19
GPU
ãƒªãƒ³ã‚¯
ã‚¢ãƒãƒ11å·ã®ã€Œæœˆé¢ç€é™¸ã¯äº‹å®Ÿã€ï¼šNVIDIAç¤¾ãŒæœ€æ–°GPUæŠ€è¡“ã§è¨¼æ˜Ž
gigs123 2014/10/01
NASA

ç§‘å¦

å¤©æ–‡

GPU
ãƒªãƒ³ã‚¯
ï¼»SIGGRAPHï¼½Pixarï¼Œã€ŒOpenSubdivã€ã«ã‚ˆã‚ŠCatmull-Clarkæ³•ã®ãƒ†ãƒƒã‚»ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³æŠ€è¡“ã‚’ã‚ªãƒ¼ãƒ—ãƒ³åŒ–ã€‚DCCãƒ„ãƒ¼ãƒ«ã‹ã‚‰ã‚²ãƒ¼ãƒ ã‚¨ãƒ³ã‚¸ãƒ³ã¾ã§ãŒPixarå“è³ªã«!?
ï¼»SIGGRAPHï¼½Pixarï¼Œã€ŒOpenSubdivã€ã«ã‚ˆã‚ŠCatmull-Clarkæ³•ã®ãƒ†ãƒƒã‚»ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³æŠ€è¡“ã‚’ã‚ªãƒ¼ãƒ—ãƒ³åŒ–ã€‚DCCãƒ„ãƒ¼ãƒ«ã‹ã‚‰ã‚²ãƒ¼ãƒ ã‚¨ãƒ³ã‚¸ãƒ³ã¾ã§ãŒPixarå“è³ªã«!? ãƒ©ã‚¤ã‚¿ãƒ¼ï¼šè¥¿å·å–„å¸ ã€ŒToy Storyã€ï¼ˆãƒˆã‚¤ã‚¹ãƒˆãƒ¼ãƒªãƒ¼ï¼‰ã‚·ãƒªãƒ¼ã‚ºãªã©ã®CGã‚¢ãƒ‹ãƒ¡ãƒ¼ã‚·ãƒ§ãƒ³åˆ¶ä½œã‚¹ã‚¿ã‚¸ã‚ªã¨ã—ã¦çŸ¥ã‚‰ã‚Œã¦ã„ã‚‹Pixar Animation Studiosï¼ˆãƒ”ã‚¯ã‚µãƒ¼ãƒ»ã‚¢ãƒ‹ãƒ¡ãƒ¼ã‚·ãƒ§ãƒ³ãƒ»ã‚¹ã‚¿ã‚¸ã‚ªï¼Œä»¥ä¸‹ï¼ŒPixarï¼‰ã ãŒï¼Œå®Ÿã¯CG æ˜ ç”»æ¥ç•Œã§åºƒãä½¿ã‚ã‚Œã¦ã„ã‚‹ãƒ¬ãƒ³ãƒ€ãƒ©ãƒ¼ã€ŒRenderManã€ã‚’ã¯ã˜ã‚ã¨ã—ãŸCGåˆ¶ä½œé–¢é€£ãƒ„ãƒ¼ãƒ«ã‚„ï¼Œãƒ™ãƒ¼ã‚¹ã¨ãªã‚‹æŠ€è¡“é–‹ç™ºã‚’è¡Œã£ã¦ã„ã‚‹ã“ã¨ã¯ã‚ã¾ã‚ŠçŸ¥ã‚‰ã‚Œã¦ã„ãªã„ã‹ã‚‚ã—ã‚Œãªã„ã€‚ ãã‚“ãªPixarã¯ï¼ŒSIGGRAPHã«æ¯Žå¹´å‡ºå±•ã—ï¼ŒåŒç¤¾ã®æŠ€è¡“è€…ã‚„ã‚¢ãƒ¼ãƒ†ã‚£ã‚¹ãƒˆãŒã•ã¾ã–ã¾ãªã‚»ãƒƒã‚·ãƒ§ãƒ³ã§è¬›æ¼”ã‚’è¡Œã£ã¦ã„ã‚‹ã€‚ä»Šå¹´ã®SIGGRAPH 2012ã«ã‚‚åŒç¤¾ã¯ä¾‹å¹´ã©ãŠã‚Šå‡ºå±•ã—ã¦ãŠã‚Šï¼Œéžå¸¸ã«ãƒ¦ãƒ‹ãƒ¼ã‚¯ã‹ã¤å¤§ããªç™º
gigs123 2012/08/14
CG

ã‚¢ãƒ‹ãƒ¡ãƒ¼ã‚·ãƒ§ãƒ³

GPU

3D
ãƒªãƒ³ã‚¯
Raspberry Pi - Wikipedia
Raspberry Piï¼ˆãƒ©ã‚ºãƒ™ãƒªãƒ¼ ãƒ‘ã‚¤ï¼‰ã¯ã€ARMãƒ—ãƒã‚»ãƒƒã‚µã‚’æè¼‰ã—ãŸã‚·ãƒ³ã‚°ãƒ«ãƒœãƒ¼ãƒ‰ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã€‚ã‚¤ã‚®ãƒªã‚¹ã®Raspberry Pi è²¡å›£ï¼ˆéžå–¶åˆ©çµ„ç¹”ï¼‰ã¨Raspberry Pi Ltdï¼ˆå–¶åˆ©çµ„ç¹”ï¼‰ã«ã‚ˆã£ã¦é–‹ç™ºã•ã‚Œã¦ã„ã‚‹ã€‚æ—¥æœ¬èªžã§ã¯ç•¥ç§°ã¨ã—ã¦ãƒ©ã‚ºãƒ‘ã‚¤ã¨ã‚‚å‘¼ã°ã‚Œã‚‹[3]ã€‚ 2012å¹´ã®ç«‹ã¡ä¸Šã’å½“åˆã¯æ•™è‚²ã§åˆ©ç”¨ã•ã‚Œã‚‹ã“ã¨ã‚’æƒ³å®šã—ã¦åˆ¶ä½œã•ã‚ŒãŸã€‚IoTãŒéš†ç››ã—ãŸ2010å¹´ä»£å¾ŒåŠä»¥é™ã¯ã€å®‰ä¾¡ã«å…¥æ‰‹ã§ãã‚‹ã‚·ãƒ³ã‚°ãƒ«ãƒœãƒ¼ãƒ‰ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã¨ã—ã¦è¶£å‘³ã‚„æ¥å‹™ï¼ˆè©¦ä½œå“ã®é–‹ç™ºï¼‰ç‰ã¨ã—ã¦ã‚‚ç”¨ã„ã‚‰ã‚Œã‚‹ã‚ˆã†ã«ãªã£ãŸã€‚ãã®å¾ŒRaspberry Pi Compute Moduleã‚’å•†å“ã«çµ„ã¿è¾¼ã‚€ç”¨é€”ã¾ã§åºƒãŒã£ã¦ã„ãã€2023å¹´ã¯å£²ä¸Šã®72%ãŒç”£æ¥ãƒ»çµ„ã¿è¾¼ã¿å¸‚å ´ã¨ãªã£ã¦ã„ã‚‹[4][5]ã€‚ Raspberry Piã¯ã€ã‹ã¤ã¦ã‚¤ã‚®ãƒªã‚¹ã§æ•™è‚²ç”¨ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿ã¨ã—ã¦æ™®åŠã—ãŸã‚¨ã‚¤ã‚³ãƒ¼ãƒ³ç¤¾ã®ã€ŒBBC Microã€ï¼ˆ1981å¹´ï¼‰ã®å†æ¥ã¨ã—ã¦
gigs123 2012/06/03
Raspberry Pi

ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ã‚¿

GPU

Linux

python
ãƒªãƒ³ã‚¯
GPUã®è¨ˆç®—èƒ½åŠ›ã‚’ãƒ–ãƒ©ã‚¦ã‚¶ã‹ã‚‰æ‰±ãˆã‚‹ã€ŒWebCLã€ç™»å ´ - Webkitå‘ã‘OpneCLå®Ÿè£… | ã‚¨ãƒ³ã‚¿ãƒ¼ãƒ—ãƒ©ã‚¤ã‚º | ãƒžã‚¤ã‚³ãƒŸã‚¸ãƒ£ãƒ¼ãƒŠãƒ«
WebCL is defining JavaScript bindings to OpenCL. Mac OS Xã®WebKitã§ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚¢ã‚¯ã‚»ãƒ©ãƒ¬ãƒ¼ã‚¿ã®æ©Ÿèƒ½ã‚’æ´»ç”¨ã—ãŸè¨ˆç®—å‡¦ç†ãŒã§ãã‚‹ã‚ˆã†ã«ãªã£ãŸã€‚SamsungãŒOpenCL 1.0ã«å¯¾å¿œã—ãŸå®Ÿè£…ã‚’ã€ŒWebCL for WebKitã€ã¨ã—ã¦Google Codeã§å…¬é–‹ã—ãŸãŸã‚ã ã€‚å®Ÿè£…ã¯BSDãƒ©ã‚¤ã‚»ãƒ³ã‚¹ã®ã‚‚ã¨ã§æä¾›ã•ã‚Œã¦ã„ã‚‹ã€‚åˆ©ç”¨ã™ã‚‹ã«ã¯NVIDIAã®ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ãƒãƒƒãƒ—ãŒæè¼‰ã•ã‚ŒãŸMacãŒå¿…è¦ã ãŒã€Mac OS Xã®ãƒ–ãƒ©ã‚¦ã‚¶ã§OpenCLãŒæ‰‹è»½ã«åˆ©ç”¨ã§ãã‚‹ã‚ˆã†ã«ãªã£ãŸã‚‚ã®ã¨ã—ã¦æ³¨ç›®ã«å€¤ã™ã‚‹ã€‚ WebCLã¯OpenCLã®å®Ÿè£…ç³»ã€‚JavaScriptã‹ã‚‰GPUã‚’è¨ˆç®—ç”¨ã®ãƒ‡ãƒã‚¤ã‚¹ã¨ã—ã¦åˆ©ç”¨ã§ãã‚‹ã€‚GPUã®æ©Ÿèƒ½ã‚’æ•°å€¤è¨ˆç®—ã«åˆ©ç”¨ã™ã‚‹ã¨ã„ã£ãŸã“ã¨ã‚’ç›®çš„ã«ã—ãŸã‚‚ã®ã§ã€CPUã®è² è·ã‚’å¢—ã‚„ã™ã“ã¨ãªãã‚ˆã‚Šå¤šãã®è¨ˆç®—ãŒè¡Œãˆã‚‹ã€‚ã‚·ãƒŸãƒ¥ãƒ¬ãƒ¼ã‚·ãƒ§ãƒ³ã‚„ç§‘å¦æŠ€è¡“è¨ˆç®—ã€
gigs123 2011/07/07
Mac OS X

WebCL

OpenCL

GPU
ãƒªãƒ³ã‚¯
[ãƒ…] Google Chrome ã§ WebGL ãŒå‹•ä½œã—ãªã„ã®ã§ --ignore-gpu-blacklist ã§å›žé¿ã™ã‚‹æ–¹æ³• (2011-04-06)
ã‚¦ãƒã® Windows XP + Google Chrome 10 ã ã¨ WebGL ã®ã‚³ãƒ¼ãƒ‰ãŒå®Ÿè¡Œã•ã‚Œãªã„ã€‚ã€‚ã€‚ æ£ç¢ºã«ã¯Chrome 10ã€Windows XPã€Nvidia 7300 GTã®çµ„ã¿åˆã‚ã›ã§Google BodyãŒå‹•ä½œã—ãªã„äº‹ãŒç™ºç«¯ã¨ã—ã¦Devç‰ˆã‚’Windows XPã«ã‚¤ãƒ³ã‚¹ãƒˆãƒ¼ãƒ«ã—ãŸå ´åˆã€ãƒ‡ãƒ•ã‚©ãƒ«ãƒˆã§WebGLãŒdisableã«ãªã£ãŸæ§˜ã§ã™ã€‚ Chrome 10ã‹ã‚‰Windows XPã§ã®WebGLã‚µãƒãƒ¼ãƒˆã‚’åœæ¢ - å¼·ç«ã§é€²ã‚ Î£ï¼ˆã‚œÐ´ã‚œlllï¼‰ã‚¬ãƒ¼ãƒ³ï¼ ã‚¦ãƒã®ã¯ Chrome version 10.0.648.204 (Official Build 79063) â€»ä»¥ä¸‹ã®æ–¹æ³•ã¯ã‚ã¾ã‚ŠãŠå‹§ã‚å‡ºæ¥ã‚‹äº‹ã§ã¯æœ‰ã‚Šã¾ã›ã‚“ã€‚è‡ªå·±è²¬ä»»ã«ã¦ã©ã†ã—ã¦ã‚‚è©¦ã—ãŸã„æ–¹ã®ã¿è©¦ã—ã¦ã¿ã¦ä¸‹ã•ã„ã€‚ å¼•æ•°ã« --ignore-gpu-blacklist ã‚’ä»˜ã‘ã¦Chromeã‚’èµ·å‹•ã™ã‚‹ã¨ã“ã®ãƒ–ãƒ©ãƒƒã‚¯ãƒª
gigs123 2011/05/19
webGL

chrome

GPU
ãƒªãƒ³ã‚¯
ã€ã‚¤ãƒ™ãƒ³ãƒˆãƒ¬ãƒãƒ¼ãƒˆã€‘ NVIDIAãŒã€ŒGPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°2010ã€ã‚’é–‹å‚¬ ï½žGPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ãŒç”Ÿã‚€æ–°ã—ã„ã‚½ãƒªãƒ¥ãƒ¼ã‚·ãƒ§ãƒ³ã‚’ç´¹ä»‹
gigs123 2010/07/20
GPU
ãƒªãƒ³ã‚¯
Accelerator V2ã‚’ä½¿ã£ãŸãƒžãƒ«ãƒã‚³ã‚¢GPU/CPUä¸Šã§ã®ä¸¦åˆ—ãƒ—ãƒã‚°ãƒ©ãƒ ã®å®Ÿè¡Œ
ã‚ãªãŸã«ã¨ã£ã¦é‡è¦ãªãƒˆãƒ”ãƒƒã‚¯ã‚„åŒåƒšã®æœ€æ–°æƒ…å ±ã‚’å…¥æ‰‹ã—ã¾ã—ã‚‡ã†æœ€æ–°ã®æ´žå¯Ÿã¨ãƒˆãƒ¬ãƒ³ãƒ‰ã«é–¢ã™ã‚‹æœ€æ–°æƒ…å ±ã‚’å³åº§ã«å—ã‘å–ã‚Šã¾ã—ã‚‡ã†ã€‚ ç¶™ç¶šçš„ãªå¦ç¿’ã®ãŸã‚ã«ã€ç„¡æ–™ã®ãƒªã‚½ãƒ¼ã‚¹ã«æ‰‹è»½ã«ã‚¢ã‚¯ã‚»ã‚¹ã—ã¾ã—ã‚‡ã†ãƒŸãƒ‹ãƒ–ãƒƒã‚¯ã€ãƒˆãƒ©ãƒ³ã‚¹ã‚¯ãƒªãƒ—ãƒˆä»˜ãå‹•ç”»ã€ãŠã‚ˆã³ãƒˆãƒ¬ãƒ¼ãƒ‹ãƒ³ã‚°æ•™æã€‚ è¨˜äº‹ã‚’ä¿å˜ã—ã¦ã€ã„ã¤ã§ã‚‚èªã‚€ã“ã¨ãŒã§ãã¾ã™è¨˜äº‹ã‚’ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ã—ã¦ã€æº–å‚™ãŒã§ããŸã‚‰ã„ã¤ã§ã‚‚èªã‚ã¾ã™ã€‚
gigs123 2009/12/24
GPU

.Net

parallel
ãƒªãƒ³ã‚¯
ã€ãƒ¬ãƒãƒ¼ãƒˆã€‘SIGGRAPH ASIA 2009 - NVIDIAã€David KirkåŸºèª¿è¬›æ¼”ã€ŒãªãœGPGPUãŒå°é ã—ãŸã‹ã€ (2) Kirkæ°ã„ã‚ãã€ŒãªãœGPGPUãŒå°é ã—ãŸã‹ã€(1) | ãƒ‘ã‚½ã‚³ãƒ³ | ãƒžã‚¤ã‚³ãƒŸã‚¸ãƒ£ãƒ¼ãƒŠãƒ«
gigs123 2009/12/22
CUDA

ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£

GPU

ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°

ä¸¦åˆ—åŒ–
ãƒªãƒ³ã‚¯
NVIDIAã€åˆã®OpenCLã‚»ãƒŸãƒŠãƒ¼ã‚’é–‹å‚¬ ï½žã‚¢ãƒƒãƒ—ãƒ«ãŒMac OSã§ã®OpenCLã‚’ãƒ‡ãƒ¢ã»ã‹
gigs123 2009/11/02
OpenCL

CUDA

Intel

Apple

GPU

PC
ãƒªãƒ³ã‚¯
ã“ã‚Œã¯å£®å¤§ãªã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°é©å‘½ã®å§‹ã¾ã‚Šã«éŽãŽãªã„
ã“ã‚Œã¯å£®å¤§ãªã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°é©å‘½ã®å§‹ã¾ã‚Šã«éŽãŽãªã„ï¼šNVIDIA GPU Techno logy Conferenceï¼ˆ1/3 ãƒšãƒ¼ã‚¸ï¼‰ NVIDIAã¯ã€2006å¹´ã«CUDAã®å–ã‚Šçµ„ã¿ã‚’ç™ºè¡¨ã—ã¦ä»¥æ¥ã€ç©æ¥µçš„ã«GPUã‚’åˆ©ç”¨ã—ãŸæ±Žç”¨ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã«å–ã‚Šçµ„ã‚“ã§ããŸã€‚ä»Šå›žç±³å›½ã§è¡Œã‚ã‚ŒãŸGTCï¼ˆGpu Techno logy Conferenceï¼‰ã§ã¯ã€ã‚½ãƒ•ãƒˆã‚¦ã‚§ã‚¢æŠ€è¡“è€…ã«ã€ä»Šå¾Œã®GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®æ–¹å‘æ€§ã‚„ã€æ–°ã—ã„æ©Ÿèƒ½ã‚’åˆ©ç”¨ã™ã‚‹ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ã®æ–¹æ³•ã«ã¤ã„ã¦â€œé›†ä¸è¬›ç¾©â€ã‚’è¡Œã£ãŸã€‚ åŒç¤¾ã®CEOã‚’å‹™ã‚ã‚‹ã‚¸ã‚§ãƒ³ãƒ»ã‚»ãƒ³ãƒ»ãƒ•ã‚¡ãƒ³æ°ã®åŸºèª¿è¬›æ¼”ã§ã¯ã€NVIDIAã®GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã«å¯¾ã™ã‚‹è€ƒãˆã«ã¤ã„ã¦ç´¹ä»‹ã•ã‚ŒãŸã€‚ãƒ•ã‚¡ãƒ³æ°ã¯ã€Œæˆ‘ã€…ã¯ã‚¼ãƒã‹ã‚‰è¨è¨ˆã—ãŸæ–°ã—ã„GPUã®ã€ŒFermiã€ã‚’è¿‘ã„å°†æ¥ã«ãƒªãƒªãƒ¼ã‚¹ã™ã‚‹ã€‚Fermiã¯GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã«æœ€é©åŒ–ã—ãŸæœ€åˆã®GPUã¨ãªã‚‹ã ã‚ã†ã€ã¨ã€æ–°ã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã‚’æŽ¡ç”¨
gigs123 2009/10/05
ãƒ†ã‚¯ãƒŽãƒã‚¸ãƒ¼

ãƒ‹ãƒ¥ãƒ¼ã‚¹

æŠ€è¡“

GPU

Fermi

ãƒ•ã‚¡ãƒ¼ãƒŸ
ãƒªãƒ³ã‚¯
NVIDIAã€æ¬¡ä¸–ä»£CUDA GPUã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã€ŒFermiã€ç™ºè¡¨ | ãƒ‘ã‚½ã‚³ãƒ³ | ãƒžã‚¤ã‚³ãƒŸã‚¸ãƒ£ãƒ¼ãƒŠãƒ«
ç±³NVIDIAã¯9æœˆ30æ—¥(ç¾åœ°æ™‚é–“)ã€ç±³ã‚«ãƒªãƒ•ã‚©ãƒ«ãƒ‹ã‚¢å·žã‚µãƒ³ãƒŽã‚¼ã§é–‹å‚¬ä¸ã®GPU Techno logy Conferenceã§ã€æ–°ã—ã„CUDA GPUã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã€ŒFermiã€(ã‚³ãƒ¼ãƒ‰ãƒãƒ¼ãƒ )ã‚’ç™ºè¡¨ã—ãŸã€‚ç¾åœ¨GeForceã€Quadroã€Teslaãªã©ã§æ§‹æˆã•ã‚Œã¦ã„ã‚‹åŒç¤¾ã®GPUãƒ•ã‚¡ãƒŸãƒªãƒ¼ã®æ¬¡ä¸–ä»£åŸºç›¤ã¨ãªã‚‹ã€‚ã€ŒFermiã‚¢ãƒ¼ã‚ãƒ†ã‚¯ãƒãƒ£ã¯ã€ä¸–ç•Œåˆã®ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ãƒ¼ã‚·ãƒ§ãƒŠãƒ«GPUã®ãŸã‚ã®åŸºç›¤ã§ã‚ã‚Šã€ã‚¼ãƒã‹ã‚‰è¨è¨ˆã•ã‚ŒãŸã€‚ã‚°ãƒ©ãƒ•ã‚£ãƒƒã‚¯ã‚¹ãŠã‚ˆã³GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã®ä¸¡æ–¹ã«é£›èºçš„ãªé€²æ©ã‚’ã‚‚ãŸã‚‰ã™ã€ã¨ã—ã¦ã„ã‚‹ã€‚ Fermiãƒ™ãƒ¼ã‚¹ã®å‹•ä½œã™ã‚‹ãƒãƒƒãƒ—ã®å˜åœ¨ã‚’ã‚¢ãƒ”ãƒ¼ãƒ«ã™ã‚‹NVIDIA CEOã®Jen-Hsun Huangæ°ã€‚é€šå¸¸ã¯ã€ã“ã®çŠ¶æ…‹ã‹ã‚‰æ•°ãƒ¶æœˆã§è£½å“å‡ºè·ã«ã“ãŽç€ã‘ã‚‹ã¨ã„ã† Fermiã¯ã€æµ®å‹•å°æ•°ç‚¹æ¼”ç®—æ¨™æº–IEEE 754-2008ã«æº–æ‹ ã—ãŸæœ€å¤§512åŸºã®CUDAã‚³ã‚¢ã‚’ã‚µãƒãƒ¼ãƒˆã™ã‚‹ã€‚å€ç²¾åº¦æ¼”ç®—ã®ãƒ”ãƒ¼ã‚¯
gigs123 2009/10/01
CUDA

GPU
ãƒªãƒ³ã‚¯
1