[B! CUDA][cuda] ang65ã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

ang65 id:ang65

CUDAã¨cudaã«é–¢ã™ã‚‹ang65ã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (38)

${{author_name}}$

{{{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}}

Pythoné«˜é€ŸåŒ–ã€€Numbaå…¥é–€ã€€ãã®4 - tkm2261's blog
ä»Šå›žã¯Numbaã®GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã«ã¤ã„ã¦èªã‚“ã§ã„ãã¾ã™ã€‚ æœ€çµ‚å›žã®äºˆå®šã§ã—ãŸãŒã€ã‚¨ãƒ³ãƒˆãƒªãŒè¶…é•·ããªã‚Šãã†ãªã®ã§ä»Šå›žã¯GPUã®ä½¿ç”¨æ–¹æ³•ã€æ¬¡å›žã«è¨ˆç®—é€Ÿåº¦ã®æ¤œè¨¼ã‚’ã—ã¦çµ‚ã‚ã‚ŠãŸã„ã¨æ€ã„ã¾ã™ã€‚ Writing CUDA-Python â€” numba 0.15.1 documentation exprimentalæ‰±ã„ãªã®ã§ã€å•†ç”¨ã§ä½¿ã‚ã‚Œã‚‹æ–¹ã¯NumbaProã®æ–¹ã‚’ã‚ªã‚¹ã‚¹ãƒ¡ã—ã¾ã™ã€‚ Numbaã®GPUã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã¯ã“ã‚Œã¾ã§ã®ãƒ‡ã‚³ãƒ¬ãƒ¼ã‚¿ã ã‘ã‚Œä½¿ãˆã‚‹Numbaã®çŠ¶æ³ã¨ã¯ç•°ãªã‚Šã€ã‹ãªã‚ŠCUDAã«é–¢ã™ã‚‹çŸ¥è˜ã‚’è¦æ±‚ã•ã‚Œã¾ã™ã€‚ ã‚‚ã—ã€å°‘ãªã„åŠ´åŠ›ã§Pythonã‚³ãƒ¼ãƒ‰ã‚’é«˜é€ŸåŒ–ã—ãŸã„æ–¹ã¯ã€ã€Žãã®ï¼“ã€ã¾ã§ã®å†…å®¹ã§ååˆ†ã¨æ€ã‚ã‚Œã¾ã™ã€‚ CUDAã«ã¤ã„ã¦ ã¨ã‚Šã‚ãˆãšã€Wikipediaã‹ã‚‰å¼•ç”¨ CUDAï¼ˆCompute Unified Device Architectureï¼šã‚¯ãƒ¼ãƒ€ï¼‰ã¨ã¯ã€NVIDIAãŒæä¾›ã™ã‚‹
ang65 2014/12/30
CUDA

python

é«˜é€ŸåŒ–
ãƒªãƒ³ã‚¯
Google Code Archive - Long-term storage for Google Code Project Hosting.
ang65 2014/09/08
CUDA
ãƒªãƒ³ã‚¯
Separate Compilation and Linking of CUDA C++ Device Code | NVIDIA Technical Blog
Managing complexity inÂ large programs requires breaking them down into components that are responsible for small, well-defined portions of the overall program. Separate compilation is an integral part of the C and C++ programming languages which allows portions of a program to be compiled into separate objects and then linked together to form an executable or library.Â Developing large and complex
ang65 2014/08/22
CUDA
ãƒªãƒ³ã‚¯
CUDA Pro Tip: Improve NVIDIA Visual Profiler Loading of Large Profiles | NVIDIA Technical Blog
ang65 2014/08/22
CUDA
ãƒªãƒ³ã‚¯
CUDA Pro Tip: nvprof is Your Handy Universal GPU Profiler | NVIDIA Technical Blog
CUDA 5 added a powerful new tool to the CUDA Toolkit: nvprof. nvprof is a command-line profiler available for Linux, Windows, and OS X. At first glance, nvprof seems to be just a GUI-less version of the graphical profiling features available in the NVIDIA Visual ProfilerÂ and NSight Eclipse edition. But nvprof is much more than that; to me, nvprof is the light-weight profiler that reaches where oth
ang65 2014/08/22
CUDA
ãƒªãƒ³ã‚¯
Adaptive Parallel Computation with CUDA Dynamic Parallelism | NVIDIA Technical Blog
Early CUDA programs had to conform to a flat, bulk parallel programming model. Programs had to perform a sequence of kernel launches, and for best performance each kernel had to expose enough parallelism to efficiently use the GPU. For applications consisting of â€œparallel forâ€ loops the bulk parallel model is notÂ too limiting, but some parallel patternsâ€”such as nested parallelismâ€”cannot be express
ang65 2014/08/10
parallel

CUDA

kernel
ãƒªãƒ³ã‚¯
MapReduce Word Count with GPUs using Pycuda
ang65 2013/11/06
CUDA

python

wordcount
ãƒªãƒ³ã‚¯
ï½¢ã‚¹ã‚«ãƒ©ãƒ¼ï½¥ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°è¬›ç¿’ä¼šï½£ã¨ï½¢ä¸¦åˆ—ãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°(MPI)è¬›ç¿’ä¼šï½£ã§é…å¸ƒã—ãŸãƒ†ã‚ã‚¹ãƒˆï½œHPC High Performance Computingï½œACCC. RIKEN
ï¼ˆ2009å¹´7æœˆ21æ—¥æ›´æ–°ï¼‰ æ³¨æ„ åŽŸç¨¿ã‚’ã‚¤ãƒ¡ãƒ¼ã‚¸ãƒ‡ãƒ¼ã‚¿ã¨ã—ã¦èªã¿è¾¼ã‚“ã§PDFãƒ•ã‚¡ã‚¤ãƒ«åŒ–ã—ãŸãŸã‚ ï½¤ ã‚µã‚¤ã‚ºãŒå¤§ãããªã£ã¦ã„ã¾ã™ã€‚ å…¨ãƒšãƒ¼ã‚¸ã‚’å«ã‚€ãƒ•ã‚¡ã‚¤ãƒ«ã¨ ï½¤ ç« (ã¾ãŸã¯ã„ãã¤ã‹ã®ç¯€)ã”ã¨ã«åˆ†å‰²ã—ãŸãƒ•ã‚¡ã‚¤ãƒ«ã®2ç¨®é¡žãŒã‚ã‚Šã¾ã™ï¼ˆå†…å®¹ã¯åŒã˜ã§ã™ï¼‰ã€‚ ãƒ†ã‚ã‚¹ãƒˆã®å†…å®¹ã¯ä»Šå¾Œä¸å®šæœŸã«åŠ ç†ä¿®æ£ã—ã¾ã™ ï½¡ ãƒ•ã‚¡ã‚¤ãƒ«åã®æœ«å°¾ã«ã¤ã„ã¦ã„ã‚‹ ï½¢20xx-xx-xx ï½£ ã¯ã€ãã®ãƒ•ã‚¡ã‚¤ãƒ«å†…ã®ãƒšãƒ¼ã‚¸ãŒæœ€å¾Œã«æ›´æ–°ã•ã‚ŒãŸæ—¥ä»˜ã‚’ç¤ºã—ã¾ã™ ï½¡ ç°¡æ˜“è£½æœ¬ã™ã‚‹éš›ã¯ ï½¤i, iii, vãƒšãƒ¼ã‚¸(vã¯(2)ã®ã¿) ï½¤ ãŠã‚ˆã³æœ¬æ–‡ã®å¥‡æ•°ãƒšãƒ¼ã‚¸(ä¾‹ãˆã°1-1,1-3, ï½¥ ï½¥ ï½¥,2-1,2-3, ï½¥ ï½¥ ï½¥)ãŒ ï½¤ è¦‹é–‹ãã—ãŸå³å´ã®ãƒšãƒ¼ã‚¸ã«ãªã‚‹ã‚ˆã†ã«ã—ã¦ä¸‹ã•ã„ ï½¡ ä¸‹è¨˜è³‡æ–™ã«é–¢ã™ã‚‹ãŠå•ã„åˆã‚ã›ã¯ ã¾ã§ã”é€£çµ¡ä¸‹ã•ã„ ï½¡ (1) ã€Œãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°æŠ€æ³•å…¥é–€ã€(æ—§ã€Œãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°æŠ€æ³•è™Žã®å·»ã€) ã€Œã‚¹ã‚«ãƒ©ãƒ¼ãƒ»ãƒãƒ¥ãƒ¼ãƒ‹ãƒ³ã‚°è¬›ç¿’ä¼šã€ ã§é…
ang65 2012/01/10
MPI

hpc

é«˜é€ŸåŒ–

PDF

è³‡æ–™

cuda

performance
ãƒªãƒ³ã‚¯
CUDAã®ã‚«ãƒ¼ãƒãƒ«é–¢æ•°ã§printfã‚’ä½¿ã† - ãã‚Œãªã‚Šã«ãƒžã‚¸ãƒ¡ãªãƒ¡ãƒ¢
ã‚«ãƒ¼ãƒãƒ«é–¢æ•°ã§ã‚‚printf()ãŒä½¿ãˆã‚‹ã‚ˆã†ã«ãªã£ãŸã®ã«å…¨ç„¶ä½¿ã£ã¦ã“ãªã‹ã£ãŸã®ã§ã€ æŠ˜è§’ãªã®ã§ãŸã‚ãã†ã¨CUDAãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ã‚¬ã‚¤ãƒ‰ã®ã‚½ãƒ¼ã‚¹ã‚’ãã®ã¾ã¾æŒã£ã¦ãã¦ã‚³ãƒ³ãƒ‘ã‚¤ãƒ«ã—ãŸã‚‰ãƒ»ãƒ»ãƒ» error: calling a host function from a __device__/__global__ function is not allowed ã£ã¦è¨€ã‚ã‚ŒãŸorz ãƒãƒ¼ã‚¸ãƒ§ãƒ³ã‚‚nvcc3.2ã§printf()ãŒä½¿ãˆã‚‹ã£ã¦æ›¸ã„ã¦ã‚ã‚‹ã€‚ GPUã‚‚Fermiã‚’ä½¿ã£ã¦ã‚‹ã®ã«ãªã‚“ã§ï¼Ÿã£ã¦ãªã£ãŸæ™‚ã®ãƒ¡ãƒ¢ã§ã™ã€‚ ç’°å¢ƒ SUSE Linux Enterprise Server 11 SP1 Tesla M2050 nvcc 3.2 è§£æ±ºç– ãŸã‚ã—ã«å‚è€ƒè³‡æ–™ã®CUDAãƒ—ãƒã‚°ãƒ©ãƒŸãƒ³ã‚°ã‚¬ã‚¤ãƒ‰ã®ä¸ã®B.14 Formatted Outputã®éƒ¨åˆ†ã«ã‚ã‚‹ã‚µãƒ³ãƒ—ãƒ«ã‚’ãã®ã¾ã¾ã‚³ãƒ³ãƒ‘ã‚¤ãƒ«ã—ã¦ã„ã‚‹ã®ã«ãƒ€ãƒ¡ã ã£ãŸã®ã§ãªã‚“
ang65 2011/04/26
CUDA

GPGPU
ãƒªãƒ³ã‚¯
memoryã‚’ã‚¤ãƒ³ã‚¯ãƒ«ãƒ¼ãƒ‰ã—ãŸã‚³ãƒ¼ãƒ‰ã‚’nvccã§ã‚³ãƒ³ãƒ‘ã‚¤ãƒ«ã—ãŸã‚‰ã‚¨ãƒ©ãƒ¼ãŒå‡ºãŸorz - ãã‚Œãªã‚Šã«ãƒžã‚¸ãƒ¡ãªãƒ¡ãƒ¢
CUDAã®ã‚³ãƒ¼ãƒ‰ã®ä¸ï¼ˆã‚«ãƒ¼ãƒãƒ«é–¢æ•°ã˜ã‚ƒãªã„ã‚ˆï¼‰ã§shared_ptrã‚’ä½¿ã†ãŸã‚ã«memoryã‚’ã‚¤ãƒ³ã‚¯ãƒ«ãƒ¼ãƒ‰ã—ãŸã‚‰ã‚¨ãƒ©ãƒ¼ãŒå‡ºã¦ã‚³ãƒ³ãƒ‘ã‚¤ãƒ«ã§ããªã‹ã£ãŸãƒ»ãƒ»ãƒ»ã€‚ ç’°å¢ƒ SUSE Linux Enterprise Server 11 SP1 nvcc 3.2 gcc 4.3.4 ã‚¨ãƒ©ãƒ¼ãƒ¡ãƒƒã‚»ãƒ¼ã‚¸ã¨ã€ã¨ã‚Šã‚ãˆãšã®è§£æ±ºæ³• ã‚³ãƒ³ãƒ‘ã‚¤ãƒ«æ™‚ã«å‡ºã¦ããŸã‚¨ãƒ©ãƒ¼ã¯ã“ã‚“ãªã® /usr/include/c++/4.3/tr1/type_traits(240): error: invalid alignment value specified by attribute ã©ã†ã—ãŸã‚‚ã®ã‹ã¨æ€ã£ãŸãŒã€å‚è€ƒè³‡æ–™ã‚’ã‚‚ã¨ã«ä¸‹ã®ã‚ˆã†ã«ã—ã¾ã—ãŸã€‚ #define __aligned__ ignored #include #undef __aligned__ ã“ã‚Œã§ã¨ã‚Šã‚ãˆãšã¯ã‚¨ãƒ©ãƒ¼ã¯å‡ºãªã„ã—ã¡ã‚ƒã‚“ã¨å‹•ã„ã¦ã‚‹ã£ã½ã„ã‘ã©ã“ã‚Œã§ãƒ›ãƒ³ãƒˆã«ã„ã„
ang65 2011/04/18
CUDA
ãƒªãƒ³ã‚¯
#1 AI ã‚«ãƒ³ãƒ•ã‚¡ãƒ¬ãƒ³ã‚¹ - GPU ãƒ†ã‚¯ãƒŽãƒã‚¸ ã‚«ãƒ³ãƒ•ã‚¡ãƒ¬ãƒ³ã‚¹ - NVIDIA
GTC is coming back to San Jose on March 17â€“21, 2025. Join thousands of developers, innovators, and business leadersÂ to experience how AI and accelerated computing are helping humanity solve our most complex challenges. From the must-see keynote by NVIDIA CEO Jensen Huang to over 900 inspiring sessions, 300+ exhibits, technical hands-on training, and tons of unique networking events, GTC is the pla
ang65 2011/02/11
GTCï¼’ï¼ï¼‘ï¼ã®å‹•ç”»ãŒè¦‹ã‚Œã‚‹

CUDA

GPGPU
ãƒªãƒ³ã‚¯
CUDAã®ã‚«ãƒ¼ãƒãƒ«é–¢æ•°ãŒã‚¿ã‚¤ãƒ ã‚¢ã‚¦ãƒˆã—ã¦ã—ã¾ã†å ´åˆã®å¯¾å‡¦ - ãã‚Œãªã‚Šã«ãƒžã‚¸ãƒ¡ãªãƒ¡ãƒ¢
å…ˆæ—¥è³¼å…¥ã—ãŸGeForce GTX 480ã‚’è©¦ãã†ã¨ã—ãŸã¨ã“ã‚ä»¥ä¸‹ã®ã‚ˆã†ãªã‚¨ãƒ©ãƒ¼ãŒå‡ºã¦ãã¾ã—ãŸã€‚ Cuda error: Kernel execution failed in file 'testbed.cu' in line 367 : the launch timed out and was terminated. åŽŸå› ã¨ã—ã¦ã¯Primary Deviceã«æŒ‡å®šã•ã‚Œã¦ã„ã‚‹GPUã§é•·æ™‚é–“ã‚«ãƒ¼ãƒãƒ«é–¢æ•°ã‚’å®Ÿè¡Œã—ã™ãŽã¦ã„ã‚‹ãŸã‚ã€‚ å®Ÿéš›ã«æœ€å¤§å®Ÿè¡Œæ™‚é–“ãŒã©ã‚Œã ã‘ã‹ã¯OSã«ã‚ˆã£ã¦ãƒãƒ©ãƒãƒ©ãªã‚ˆã†ã§ã‚«ãƒ¼ãƒãƒ«é–¢æ•°ã®å®Ÿè¡Œæ™‚é–“ãŒé•·æ™‚é–“ã®å ´åˆã¯æ³¨æ„ãŒå¿…è¦ã€‚ å¯¾å‡¦æ³•ã¨ã—ã¯ãã‚‚ãã‚‚é•·æ™‚é–“å®Ÿè¡Œã›ãšã«ç´°ã‹ãä½•åº¦ã‚‚å®Ÿè¡Œã™ã‚‹ãªã©è€ƒãˆã‚‰ã‚Œã¾ã™ãŒã€ä½•åº¦ã‚‚å®Ÿè¡Œã™ã‚‹ã¨ã•ã¾ã–ã¾ãªã‚ªãƒ¼ãƒãƒ¼ãƒ˜ãƒƒãƒ‰ãŒã‹ã‹ã‚‹ãŸã‚ãªã‚‹ã¹ã1åº¦ã«é•·æ™‚é–“å®Ÿè¡Œã—ãŸã„å ´åˆãŒã‚ã‚Šã¾ã™ã€‚ ä»Šå›žã®ç§ã®ã‚±ãƒ¼ã‚¹ãŒãã‚Œã§ã€è§£æ±ºã—ãŸã¨ãã®ãƒ¡ãƒ¢ã€‚ XWindowã‚’æ¢ã‚ã¦ã—ã¾ã†ã¨ã„ã†
ang65 2010/08/31
CUDA

GPU
ãƒªãƒ³ã‚¯
Robotics and Embedded Systems
ang65 2010/07/22
eclipse

CUDA
ãƒªãƒ³ã‚¯
Quickstart: CUDA using Bayreuth University CUDA Toolchain for Eclipse
ang65 2010/07/22
eclipse

CUDA
ãƒªãƒ³ã‚¯
GPUãƒãƒ£ãƒ¬ãƒ³ã‚¸2010ã®ãƒ¡ãƒ¢ - ãã‚Œãªã‚Šã«ãƒžã‚¸ãƒ¡ãªãƒ¡ãƒ¢
ã¡ã‚‡ã£ã¨å‰ã§ã™ãŒ5æœˆ27æ—¥ã«SACSIS2010ã«è¡Œã£ã¦GPUãƒãƒ£ãƒ¬ãƒ³ã‚¸2010ã®è‡ªç”±èª²é¡Œã«å¿œå‹Ÿã—ã¦å„ªå‹ã—ãŸã¨ãã®å†…å®¹ã‚’ç™ºè¡¨ã—ã¦ãã¾ã—ãŸã€‚ GPUãƒãƒ£ãƒ¬ãƒ³ã‚¸ã«ã¤ã„ã¦ã¯â†“ GPU Challenge 2010 ã‚ã¨ãƒžã‚¤ã‚³ãƒŸã§ã‚‚å–ã‚Šä¸Šã’ã‚‰ã‚Œã¦ã„ãŸã®ã§ãã¡ã‚‰ã‚‚ã©ã†ãžã€‚ NVIDIAã€å¦ç”Ÿå‚åŠ ã‚³ãƒ³ãƒ†ã‚¹ãƒˆã€ŒGPUãƒãƒ£ãƒ¬ãƒ³ã‚¸2010ã€ã®è¡¨å½°å¼ | ãƒ‘ã‚½ã‚³ãƒ³ | ãƒžã‚¤ã‚³ãƒŸã‚¸ãƒ£ãƒ¼ãƒŠãƒ« å¦ç”Ÿå‚åŠ ã‚³ãƒ³ãƒ†ã‚¹ãƒˆã£ã¦æ›¸ã„ã¦ã‚ã‚‹ã‘ã©è‡ªç”±èª²é¡Œã¯åŠ©æ•™æŽˆã®äººã¨ã‹ã‚‚okãªã®ã§å¦ç”Ÿã ã‘ã˜ã‚ƒãªã„ã‚“ã§ã™ã‘ã©ãï½— å¿œå‹Ÿã—ãŸè«–æ–‡ï¼Ÿã¯ä»¥ä¸‹ã‹ã‚‰è¦‹ã‚‹ã“ã¨ãŒã§ãã¾ã™ã®ã§ã‚ˆã‚ã—ã‘ã‚Œã°ã”è¦§ãã ã•ã„ã€‚ GPUã«ã‚ˆã‚‹DNAæ–ç‰‡é…åˆ—ã®é«˜é€Ÿãƒžãƒƒãƒ”ãƒ³ã‚°(pdf) ä»Šå›žã¯è¦å®šèª²é¡Œã®ãƒ—ãƒã‚°ãƒ©ãƒ ã‚„è‡ªç”±èª²é¡Œã®ãƒ¬ãƒãƒ¼ãƒˆã¯ã™ã¹ã¦å…¬å¼ãƒšãƒ¼ã‚¸ã«ã¦å…¬é–‹ã•ã‚Œã¦ã„ã¾ã™ã®ã§èˆˆå‘³ã®ã‚ã‚‹æ–¹ã¯ãã¡ã‚‰ã‚‚ã©ã†ãžã€‚ ãƒ›ãƒ³ãƒˆã¯ã²ã£ãã‚Šã¨å‚åŠ ã—ã¦ãŸè¦å®šèª²é¡Œã®é«˜é€ŸåŒ–ã«ã¤ã„ã¦ã‚ã‚Œã“ã‚Œæ›¸ã“ã†ã¨æ€ã£ãŸã‘
ang65 2010/06/15
CUDA

GPU

hpc
ãƒªãƒ³ã‚¯
NVIDIA Blog
ang65 2010/05/18
C++supportã«ãªã£ã¦ã‚‹ï½—

GPGPU

CUDA
ãƒªãƒ³ã‚¯
Home - Fixstars Corporation
ang65 2010/04/26
Eclipseã®CUDAç”¨ã®ãƒ—ãƒ©ã‚°ã‚¤ãƒ³ãŒã‚ã‚‹ã‚‰ã—ã„ãƒ»ãƒ»ãƒ»ãã‚Œã ã‘ã»ã—ã„ã‚“ã§ã™ãŒ

CUDA

Eclipse

GPU
ãƒªãƒ³ã‚¯
Page Not Found | NVIDIA
Creatives/Designers Data Scientists Developers Executives Gamers ISVs IT Professionals Researchers Roboticists Startups
ang65 2010/03/06
ï¼‘GPUã ã¨ã‚‚ã¨ã®ï¼–å€ç¨‹åº¦ã£ã½ã„

CUDA

BLAST
ãƒªãƒ³ã‚¯
[PDF] CUDAã¨GPUã‚’ç”¨ã„ãŸSQLiteã®é«˜é€ŸåŒ–
ang65 2010/02/25
GPU

sqlite

CUDA

é«˜é€ŸåŒ–

gpgpu

database
ãƒªãƒ³ã‚¯
High-throughput sequence alignment using Graphics Processing Units - PMC
ang65 2010/02/12
CUDAã§Sufiix treeä¸Šã‚’æŽ¢ç´¢

CUDA

suffix tree

matching

è«–æ–‡
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸