{"id":31934,"date":"2023-05-23T13:08:53","date_gmt":"2023-05-23T05:08:53","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=31934"},"modified":"2023-05-23T13:08:55","modified_gmt":"2023-05-23T05:08:55","slug":"%e5%8a%9f%e7%8e%87%e9%99%8d%e4%bd%8e-%e6%9b%b4%e7%a7%91%e5%ad%b8%e3%80%81%e6%9b%b4%e5%b0%91%e8%83%bd%e8%80%97%ef%bc%8c%e5%85%a8%e7%90%83%e6%9c%80%e5%a4%a7%e8%b6%85%e7%b4%9a%e9%9b%bb%e8%85%a6","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=31934","title":{"rendered":"\u529f\u7387\u964d\u4f4e\u2014\u66f4\u79d1\u5b78\u3001\u66f4\u5c11\u80fd\u8017\uff0c\u5168\u7403\u6700\u5927\u8d85\u7d1a\u96fb\u8166\u4e4b\u4e00\u7684\u61c9\u7528\u8b49\u660eNVIDIA GPU\u80fd\u6548\u9818\u5148"},"content":{"rendered":"\n
\"\"<\/figure>\n\n\n\n

<\/p>\n\n\n\n

\u5927\u5bb6\u8a8d\u540c\uff1a\u52a0\u901f\u904b\u7b97<\/a>\u662f\u7bc0\u80fd\u7684\u904b\u7b97\u3002<\/p>\n\n\n\n

\u570b\u5bb6\u80fd\u6e90\u7814\u7a76\u79d1\u5b78\u904b\u7b97\u4e2d\u5fc3\uff08NERSC<\/a>\uff09\u662f\u7f8e\u570b\u80fd\u6e90\u90e8\u958b\u653e\u79d1\u5b78\u7684\u4e3b\u8981\u6a5f\u69cb\uff0c\u6e2c\u91cf\u4e86\u5176\u56db\u500b\u4e3b\u8981\u9ad8\u6548\u80fd\u904b\u7b97\u548c AI \u61c9\u7528\u7a0b\u5f0f\u7684\u7d50\u679c\u3002<\/p>\n\n\n\n

\u4ed6\u5011\u6e2c\u8a66\u4e86\u5728 Perlmutter<\/a>\uff08\u4f7f\u7528 NVIDIA GPU \u7684\u5168\u7403\u6700\u5927\u8d85\u7d1a\u96fb\u8166\u4e4b\u4e00\uff09\u4e0a\u50c5\u4f7f\u7528 CPU \u548c\u4f7f\u7528 GPU \u52a0\u901f\u7bc0\u9ede\u6642\uff0c\u61c9\u7528\u7a0b\u5f0f\u904b\u884c\u901f\u5ea6\u4ee5\u53ca\u80fd\u8017\u7684\u60c5\u6cc1\u3002<\/p>\n\n\n\n

\u7d50\u679c\u5f88\u660e\u986f\u3002\u4f7f\u7528 NVIDIA A100 Tensor Core GPU<\/a> \u52a0\u901f\uff0c\u80fd\u6e90\u6548\u7387\u5e73\u5747\u63d0\u9ad85\u500d\u3002\u4e00\u500b\u5929\u6c23\u9810\u5831\u61c9\u7528\u7a0b\u5f0f\u6027\u80fd\u63d0\u5347\u4e86 9.8 \u500d\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

GPU<\/strong>\u7bc0\u7701\u767e\u842c\u74e6\u7279<\/strong><\/strong><\/p>\n\n\n\n

\u5728\u4e00\u53f0\u642d\u8f09\u56db\u500b A100 GPU \u7684\u4f3a\u670d\u5668\u4e0a\uff0cNERSC \u5be6\u73fe\u4e86\u65bc\u96d9\u63d2\u69fd x86 \u4f3a\u670d\u5668\u9ad8\u9054 12 \u500d\u7684\u52a0\u901f\u3002<\/p>\n\n\n\n

\u9019\u610f\u5473\u8457\u5728\u76f8\u540c\u6027\u80fd\u6c34\u5e73\u4e0b\uff0cGPU \u52a0\u901f\u7cfb\u7d71\u6bcf\u500b\u6708\u5c07\u6bd4\u50c5\u4f7f\u7528 CPU \u7684\u7cfb\u7d71\u7bc0\u7701 588 \u5146\u74e6\u6642\u7684\u80fd\u6e90\u6d88\u8017\u3002\u5982\u679c\u5728\u4e00\u500b\u56db\u8def NVIDIA A100 \u96f2\u7aef\u57f7\u884c\u500b\u9ad4\u4e0a\u904b\u884c\u76f8\u540c\u5de5\u4f5c\u8ca0\u8f09\u4e00\u500b\u6708\uff0c\u76f8\u8f03\u65bc\u50c5\u4f7f\u7528 CPU \u7684\u57f7\u884c\u500b\u9ad4\uff0c\u7814\u7a76\u4eba\u54e1\u53ef\u4ee5\u7bc0\u7701\u8d85\u904e 400 \u842c\u7f8e\u5143\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

\u8861\u91cf\u5be6\u969b\u61c9\u7528<\/strong><\/strong><\/p>\n\n\n\n

\u9019\u4e9b\u7d50\u679c\u610f\u7fa9\u975e\u5e38\u91cd\u5927\uff0c\u56e0\u70ba\u5b83\u5011\u57fa\u65bc\u73fe\u5be6\u61c9\u7528\u7684\u6e2c\u91cf\uff0c\u800c\u4e0d\u662f\u5408\u6210\u7684\u57fa\u6e96\u3002\u9019\u4e9b\u9032\u5c55\u610f\u5473\u8457\uff0c8000 \u591a\u540d\u79d1\u5b78\u5bb6\u4f7f\u7528 Perlmutter \u80fd\u5920\u61c9\u5c0d\u66f4\u5927\u7684\u6311\u6230\uff0c\u70ba\u66f4\u591a\u7a81\u7834\u958b\u555f\u5927\u9580\u3002\u5728 Perlmutter \u4e0a\u4f7f\u7528\u8d85\u904e 7100 \u500b A100 \u7e6a\u5716\u8655\u7406\u5668\uff0c\u6709\u8a31\u591a\u61c9\u7528\u6848\u4f8b\uff0c\u79d1\u5b78\u5bb6\u6b63\u5728\u63a2\u7d22\u4e9e\u539f\u5b50\u4ea4\u4e92\u4f5c\u7528\uff0c\u4ee5\u5c0b\u627e\u65b0\u7684\u7da0\u8272\u80fd\u6e90\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

\u5168\u9762\u63a8\u9032\u79d1\u5b78\u767c\u5c55<\/strong><\/strong><\/p>\n\n\n\n

NERSC \u6240\u6e2c\u8a66\u7684\u61c9\u7528\u9818\u57df\u6db5\u84cb\u4e86\u5206\u5b50\u52d5\u529b\u5b78\u3001\u6750\u6599\u79d1\u5b78\u548c\u5929\u6c23\u9810\u6e2c\u7b49\u591a\u500b\u9818\u57df\u3002\u4f8b\u5982\uff0cMILC \u6a21\u64ec\u4e86\u5728\u539f\u5b50\u4e2d\u4f7f\u7c92\u5b50\u4fdd\u6301\u5728\u4e00\u8d77\u7684\u57fa\u672c\u529b\u91cf\u3002\u5b83\u88ab\u7528\u65bc\u767c\u5c55\u91cf\u5b50\u904b\u7b97\u3001\u7814\u7a76\u6697\u7269\u8cea\u548c\u5c0b\u627e\u5b87\u5b99\u7684\u8d77\u6e90\u3002BerkeleyGW \u5247\u6709\u52a9\u65bc\u6a21\u64ec\u548c\u9810\u6e2c\u6750\u6599\u548c\u5948\u7c73\u7d50\u69cb\u7684\u5149\u5b78\u6027\u8cea\uff0c\u9019\u662f\u958b\u767c\u66f4\u9ad8\u6548\u96fb\u6c60\u548c\u96fb\u5b50\u8a2d\u5099\u7684\u95dc\u9375\u6b65\u9a5f\u3002<\/p>\n\n\n\n

<\/p>\n\n\n

\n
\"\"
\u25b2NERSC\u61c9\u7528\u7a0b\u5f0f\u53ef\u900f\u904e\u52a0\u901f\u904b\u7b97\u7372\u5f97\u6548\u7387\u63d0\u5347<\/figcaption><\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

EXAALT\u4f7f\u7528A100 GPU\u53d6\u5f97\u4e868.5\u500d\u7684\u6548\u80fd\u63d0\u5347\uff0c\u8a72\u61c9\u7528\u7a0b\u5f0f\u89e3\u6c7a\u4e86\u5206\u5b50\u52d5\u529b\u5b78\u4e2d\u7684\u4e00\u500b\u57fa\u672c\u6311\u6230\u3002\u5b83\u8b93\u7814\u7a76\u4eba\u54e1\u53ef\u4ee5\u6a21\u64ec\u539f\u5b50\u904b\u52d5\u7684\u77ed\u7247\uff0c\u800c\u4e0d\u662f\u5176\u4ed6\u5de5\u5177\u63d0\u4f9b\u7684\u5feb\u7167\u5e8f\u5217\u3002<\/p>\n\n\n\n

\u6e2c\u8a66\u7684\u7b2c\u56db\u500b\u61c9\u7528\u7a0b\u5f0f\u662fDeepCAM\uff0c\u7528\u65bc\u5728\u6c23\u5019\u6578\u64da\u4e2d\u6aa2\u6e2c\u98b6\u98a8\u548c\u5927\u6c23\u6cb3\u6d41\u3002\u7576\u4f7f\u7528A100 GPU\u9032\u884c\u52a0\u901f\u6642\uff0c\u5b83\u7684\u80fd\u6e90\u6548\u7387\u63d0\u5347\u4e869.8\u500d\u3002<\/p>\n\n\n

\n
\"\"
\u25b2\u57fa\u65bc HPC \u548c AI \u61c9\u7528\u7a0b\u5e8f\u7684\u7d44\u5408\uff0c\u6574\u9ad4 5\u500d\u52a0\u901f<\/figcaption><\/figure><\/div>\n\n\n

<\/p>\n\n\n\n

\u52a0\u901f\u904b\u7b97\u7bc0\u7701\u66f4\u591a\u80fd\u6e90<\/strong><\/strong><\/p>\n\n\n\n

NERSC\u7684\u7d50\u679c\u8207\u65e9\u671f\u91dd\u5c0d\u52a0\u901f\u904b\u7b97\u6f5b\u5728\u7bc0\u80fd\u7684\u8a08\u7b97\u76f8\u547c\u61c9\u3002\u4f8b\u5982\uff0c\u5728NVIDIA\u7684\u53e6\u4e00\u500b\u7368\u7acb\u5206\u6790\u4e2d\uff0cGPU\u5728\u4eba\u5de5\u667a\u6167\u63a8\u7406\u65b9\u9762\u7684\u80fd\u6e90\u6548\u7387\u6bd4CPU\u9ad842\u500d\u3002<\/p>\n\n\n\n

\u9019\u610f\u5473\u8457\u5c07\u5168\u7403\u50c5\u4f7f\u7528CPU\u904b\u884c\u4eba\u5de5\u667a\u6167\u7684\u4f3a\u670d\u5668\u8f49\u63db\u70baGPU\u52a0\u901f\u7cfb\u7d71\uff0c\u6bcf\u5e74\u53ef\u7bc0\u7701\u9ad8\u905410\u5146\u74e6\u6642\u7684\u80fd\u6e90\u3002\u9019\u76f8\u7576\u65bc\u7bc0\u7701\u4e86140\u842c\u500b\u5bb6\u5ead\u4e00\u5e74\u7684\u80fd\u6e90\u6d88\u8017\u91cf\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n

\u52a0\u901f\u4f01\u696d\u767c\u5c55<\/strong><\/strong><\/p>\n\n\n\n

\u60a8\u4e0d\u5fc5\u662f\u4e00\u4f4d\u79d1\u5b78\u5bb6\uff0c\u4e5f\u80fd\u900f\u904e\u52a0\u901f\u904b\u7b97\u63d0\u5347\u80fd\u6e90\u6548\u7387\u3002<\/p>\n\n\n\n

\u88fd\u85e5\u516c\u53f8\u6b63\u5728\u4f7f\u7528GPU\u52a0\u901f\u7684\u6a21\u64ec\u548c\u4eba\u5de5\u667a\u6167\u4f86\u52a0\u901f\u85e5\u7269\u7814\u767c<\/a>\u904e\u7a0b\u3002\u50cfBMW\u96c6\u5718\u9019\u6a23\u7684\u6c7d\u8eca\u88fd\u9020\u5546\u6b63\u5728\u4f7f\u7528\u5b83\u4f86\u70ba\u6574\u500b\u5de5\u5ee0\u5efa\u6a21<\/a>\u3002<\/p>\n\n\n\n

NVIDIA\u5275\u8fa6\u4eba\u66a8\u57f7\u884c\u9577\u9ec3\u4ec1\u52f3\u6240\u8aaa\u7684\u300c\u5de5\u696d\u9ad8\u6548\u80fd\u904b\u7b97\u9769\u547d<\/a>\u300d\u53d7\u52a0\u901f\u904b\u7b97\u548c\u4eba\u5de5\u667a\u6167\u7684\u63a8\u52d5\uff0c\u800c\u9019\u4e9b\u4f01\u696d\u6b63\u662f\u8655\u65bc\u6b64\u9769\u547d\u524d\u6cbf\u4e14\u4f01\u696d\u9084\u6301\u7e8c\u589e\u52a0\u4e2d\u3002<\/p>\n\n\n\n

<\/p>\n\n\n\n


\ud83d\udfe6\u6b63\u5b97Cherry MX Ergo Clear\u7389\u8ef8\u9375\u76e4\u5927\u65b9\u9001!!\u60f3\u514d\u8cbb\u5165\u5751\u5feb\u4f86\u5206\u4eab\u62bd\u597d\u5eb7<\/a>!!
\ud83d\udfe6\u73fe\u5728\u5c31\u52a0\u5165 ioioTIMES \u81c9\u66f8\u7c89\u7d72\u5718<\/a> \u66f4\u591a\u4e92\u52d5\u3001\u66f4\u591a\u597d\u5eb7\u650f\u62b5\u52a0!!<\/strong>
\ud83d\udfe9\u6211\u5011\u6709LINE TODAY<\/a>\u983b\u9053\u4e86\uff0c\u5feb\u4f86\u8ffd\u8e2a\u6211\u5011\u5427!!–\u6700\u65b0\u79d1\u6280\u65b0\u805e \u76e1\u5728\u4f60\u624b<\/strong><\/h4>\n\n\n\n

<\/p>\n","protected":false},"excerpt":{"rendered":"

\u5927\u5bb6\u8a8d\u540c\uff1a\u52a0\u901f\u904b\u7b97\u662f\u7bc0\u80fd\u7684\u904b\u7b97\u3002 \u570b<\/p>\n","protected":false},"author":3,"featured_media":31942,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[5983,5982,580,5980,71,5981,5979,3227],"class_list":["post-31934","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-a100-gpu","tag-exaalt","tag-focus","tag-nersc","tag-nvidia","tag-perlmutter","tag-5979","tag-3227"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/31934"}],"collection":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=31934"}],"version-history":[{"count":3,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/31934\/revisions"}],"predecessor-version":[{"id":31947,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/31934\/revisions\/31947"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/31942"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=31934"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=31934"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=31934"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}