{"id":35906,"date":"2023-06-28T14:49:31","date_gmt":"2023-06-28T06:49:31","guid":{"rendered":"https:\/\/www.ioiotimes.com\/?p=35906"},"modified":"2023-06-28T14:50:25","modified_gmt":"2023-06-28T06:50:25","slug":"nvidia-h100-gpu%e7%82%ba%e9%a6%96%e6%ac%a1%e4%ba%ae%e7%9b%b8mlperf%e5%9f%ba%e6%ba%96%e6%b8%ac%e8%a9%a6%e4%b8%ad%e7%9a%84%e7%94%9f%e6%88%90%e5%bc%8fai%e8%a8%ad%e7%ab%8b%e6%a8%99%e6%ba%96","status":"publish","type":"post","link":"https:\/\/www.ioiotimes.com\/?p=35906","title":{"rendered":"NVIDIA H100 GPU\u70ba\u9996\u6b21\u4eae\u76f8MLPerf\u57fa\u6e96\u6e2c\u8a66\u4e2d\u7684\u751f\u6210\u5f0fAI\u8a2d\u7acb\u6a19\u6e96"},"content":{"rendered":"\n<h3 class=\"wp-block-heading\">\u5728\u5168\u65b0\u696d\u754c\u6a19\u6e96\u57fa\u6e96\u6e2c\u8a66\u4e2d\uff0c\u96f2\u7aef\u670d\u52d9\u4f9b\u61c9\u5546CoreWeave\u7684\u4e00\u500b\u75313,584\u500bH100 GPU\u7d44\u6210\u7684\u96c6\u7fa4\u50c5\u7528\u4e8611\u5206\u9418\u5c31\u8a13\u7df4\u5b8c\u6210\u4e00\u500b\u57fa\u65bcGPT-3\u7684\u9f90\u5927\u6a21\u578b<\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"1200\" height=\"675\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia01.png\" alt=\"\" class=\"wp-image-35922\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia01.png 1200w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia01-300x169.png 300w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia01-1024x576.png 1024w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia01-768x432.png 768w\" sizes=\"(max-width: 1200px) 100vw, 1200px\" \/><figcaption class=\"wp-element-caption\">\u25b2NVIDIA H100 GPU\u70ba\u9996\u6b21\u4eae\u76f8MLPerf\u57fa\u6e96\u6e2c\u8a66\u4e2d\u7684\u751f\u6210\u5f0fAI\u8a2d\u7acb\u6a19\u6e96<\/figcaption><\/figure>\n\n\n\n<p><\/p>\n\n\n\n<p>\u9818\u5148\u7684\u4f7f\u7528\u8005\u548c\u696d\u754c\u6a19\u6e96\u57fa\u6e96\u6e2c\u8a66\u4e00\u81f4\u540c\u610f\uff1a<a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/h100\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA H100 Tensor Core GPUs<\/a> \u63d0\u4f9b\u4e86\u6700\u4f73\u7684\u4eba\u5de5\u667a\u6167\u6548\u80fd\uff0c\u7279\u5225\u662f\u5728\u652f\u63f4<a href=\"https:\/\/www.nvidia.com\/en-us\/ai-data-science\/generative-ai\" target=\"_blank\" rel=\"noreferrer noopener\">\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167<\/a>\u7684\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff08<a href=\"https:\/\/blogs.nvidia.com\/blog\/2023\/01\/26\/what-are-large-language-models-used-for\/\" target=\"_blank\" rel=\"noreferrer noopener\">LLM<\/a>\uff09\u65b9\u9762\u3002<\/p>\n\n\n\n<p>\u5728\u4eca\u5929\u767c\u5e03\u7684\u6700\u65b0 MLPerf \u8a13\u7df4\u57fa\u6e96\u6e2c\u8a66\u4e2d\uff0cH100 GPU \u5728\u6240\u6709\u516b\u9805\u6e2c\u8a66\u4e2d\u5275\u4e0b\u4e86\u65b0\u7d00\u9304\uff0c\u5728\u91dd\u5c0d\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u7684\u65b0\u00a0MLPerf\u00a0\u6e2c\u8a66\u4e2d\u8868\u73fe\u51fa\u8272\u3002\u9019\u7a2e\u5353\u8d8a\u7684\u6027\u80fd\u80fd\u5920\u5728\u5927\u578b\u4f3a\u670d\u5668\u4e0a\u4ee5\u55ae\u500b\u52a0\u901f\u5668\u548c\u591a\u53f0 GPU \u4f3a\u670d\u5668\u7684\u65b9\u5f0f\u5c55\u73fe\u3002<\/p>\n\n\n\n<p>\u8209\u4f8b\u4f86\u8aaa\uff0c\u5c08\u9580\u5f9e\u4e8b\u00a0GPU\u00a0\u52a0\u901f\u5de5\u4f5c\u8ca0\u8f09\u7684\u96f2\u7aef\u670d\u52d9\u4f9b\u61c9\u5546\u00a0<a href=\"https:\/\/www.coreweave.com\/\" target=\"_blank\" rel=\"noreferrer noopener\">CoreWeave<\/a>\u8207\u65b0\u5275\u516c\u53f8\u00a0Inflection AI \u5171\u540c\u958b\u767c\u4e26\u7531\u00a0CoreWeave \u8ca0\u8cac\u71df\u904b\u7684\u4e00\u500b\u53ef\u5546\u7528\u7684 3,584 \u500b H100 GPU <a href=\"https:\/\/blogs.nvidia.com\/blog\/2021\/03\/05\/what-is-a-cluster-pod\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u96c6\u7fa4<\/a>\uff0c\u6b64\u7cfb\u7d71\u5728\u4e0d\u5230 11 \u5206\u9418\u7684\u6642\u9593\u5167\u5b8c\u6210\u4e86\u5927\u898f\u6a21\u7684 GPT-3 \u8a13\u7df4\u57fa\u6e96\u6e2c\u8a66\u3002<\/p>\n\n\n\n<p>CoreWeave \u806f\u5408\u5275\u8fa6\u4eba\u66a8\u6280\u8853\u9577 Brian Venturo \u8868\u793a\uff1a\u00a0\u300c\u7531\u65bc\u6211\u5011\u5728\u5feb\u901f\u3001\u4f4e\u5ef6\u9072\u00a0InfiniBand\u00a0\u7db2\u8def\u4e0a\u904b\u884c\u6578\u5343\u500b\u00a0H100 GPU\uff0c\u6211\u5011\u7684\u5ba2\u6236\u5982\u4eca\u80fd\u5927\u898f\u6a21\u5730\u69cb\u5efa\u6700\u5148\u9032\u7684\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u548c\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u3002\u6211\u5011\u8207 NVIDIA \u4e00\u540c\u63d0\u4ea4\u7684 MLPerf \u6210\u679c\u6e05\u695a\u986f\u793a\u4e86\u6211\u5011\u7684\u5ba2\u6236\u4eab\u53d7\u5230\u7684\u5353\u8d8a\u6548\u80fd\u3002\u300d<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u7576\u4eca\u6700\u9ad8\u6548\u80fd<\/strong><strong><\/strong><\/p>\n\n\n\n<p>Inflection AI \u5229\u7528\u9019\u7a2e\u6548\u80fd\u70ba\u5176\u7b2c\u4e00\u500b\u540d\u70ba <a href=\"http:\/\/heypi.com\/\" target=\"_blank\" rel=\"noreferrer noopener\">Pi<\/a> \u7684\u500b\u4eba\u4eba\u5de5\u667a\u6167\u6253\u9020\u5148\u9032\u5927\u578b\u8a9e\u8a00\u6a21\u578b\uff0c\u8a72\u516c\u53f8\u5c07\u626e\u6f14\u4e00\u500b\u4eba\u5de5\u667a\u6167\u5de5\u4f5c\u5ba4\uff0c\u6253\u9020\u8b93\u4f7f\u7528\u8005\u53ef\u4ee5\u900f\u904e\u7c21\u55ae\u3001\u81ea\u7136\u65b9\u5f0f\u8207\u4e4b\u4e92\u52d5\u7684\u500b\u4eba\u4eba\u5de5\u667a\u6167\u3002<\/p>\n\n\n\n<p>Inflection AI&nbsp;\u57f7\u884c\u9577&nbsp;Mustafa Suleyman&nbsp;\u8868\u793a\uff1a\u300c\u5982\u4eca\uff0c\u4efb\u4f55\u4eba\u90fd\u53ef\u4ee5\u9ad4\u9a57\u57fa\u65bc\u6211\u5011\u6700\u5148\u9032\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u7684\u500b\u4eba\u4eba\u5de5\u667a\u6167\u7684\u5f37\u5927\u529f\u80fd\uff0c\u8a72\u6a21\u578b\u662f\u5728&nbsp;CoreWeave&nbsp;\u5f37\u5927\u7684&nbsp;H100 GPU&nbsp;\u7db2\u8def\u4e0a\u9032\u884c\u8a13\u7df4\u7684\u3002\u300d<\/p>\n\n\n\n<p>Inflection AI \u5728 2022 \u5e74\u521d\u7531 DeepMind \u7684 Mustafa \u548c Kar\u00e9n Simonyan \u4ee5\u53ca Reid Hoffman \u5171\u540c\u5275\u7acb\uff0c\u76fc\u900f\u904e\u8207 CoreWeave \u5408\u4f5c\uff0c\u4f7f\u7528 NVIDIA \u7684 GPU \u5efa\u7acb\u5168\u7403\u6700\u5927\u7684\u904b\u7b97\u96c6\u7fa4\u4e4b\u4e00\u3002<\/p>\n\n\n\n<p><strong>\u6e2c\u8a66\u7d50\u679c<\/strong><strong><\/strong><\/p>\n\n\n\n<p>\u6e2c\u8a66\u7d50\u679c\u9019\u4e9b\u4f7f\u7528\u8005\u7d93\u9a57\u53cd\u6620\u4e86<a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/resources\/mlperf-benchmarks\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u4eca\u5929\u5ba3\u5e03<\/a>\u7684 MLPerf \u57fa\u6e96\u6e2c\u8a66\u4e2d\u5c55\u793a\u7684\u6027\u80fd\u3002<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"800\" height=\"450\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia02.png\" alt=\"\" class=\"wp-image-35925\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia02.png 800w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia02-300x169.png 300w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia02-768x432.png 768w\" sizes=\"(max-width: 800px) 100vw, 800px\" \/><\/figure><\/div>\n\n\n<p><\/p>\n\n\n\n<p>H100 GPU \u5728\u6bcf\u500b\u57fa\u6e96\u6e2c\u8a66\u4e2d\u90fd\u63d0\u4f9b\u4e86\u6700\u9ad8\u7684\u6548\u80fd\uff0c\u5305\u62ec\u5927\u578b\u8a9e\u8a00\u6a21\u578b\u3001\u63a8\u85a6\u7cfb\u7d71\u3001\u96fb\u8166\u8996\u89ba\u3001\u91ab\u5b78\u5f71\u50cf\u548c\u8a9e\u97f3\u8b58\u5225\u3002\u5b83\u5011\u662f\u552f\u4e00\u53ef\u4ee5\u57f7\u884c\u6240\u6709\u516b\u9805\u6e2c\u8a66\u7684\u6676\u7247\uff0c\u5c55\u793a\u51fa\u00a0NVIDIA AI\u00a0\u5e73\u53f0\u7684\u591a\u529f\u80fd\u6027\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u5353\u8d8a\u7684\u5927\u898f\u6a21\u57f7\u884c\u80fd\u529b<\/strong><strong><\/strong><\/p>\n\n\n\n<p>\u8a13\u7df4\u901a\u5e38\u662f\u7531\u591a\u500b GPU \u5354\u540c\u5de5\u4f5c\u7684\u5927\u898f\u6a21\u904b\u884c\u4f5c\u696d\u3002\u5728\u6bcf\u500b MLPerf \u6e2c\u8a66\u4e2d\uff0cH100 GPU \u5728 AI \u8a13\u7df4\u65b9\u9762\u5275\u4e0b\u4e86\u65b0\u7684\u5927\u898f\u6a21\u6548\u80fd\u7d00\u9304\u3002<\/p>\n\n\n\n<p>\u672c\u6b21\u63d0\u4ea4\u7684\u6e2c\u8a66\u7d50\u679c\u5305\u62ec\u6578\u767e\u500b\u5230\u6578\u5343\u500b H100 GPU\uff0c\u5c55\u73fe\u51fa\u6574\u500b\u6280\u8853\u5806\u758a\u7684\u512a\u5316\u6210\u679c\uff0c\u4e26\u5728\u56b4\u82db\u7684 LLM \u6e2c\u8a66\u4e2d\u5be6\u73fe\u4e86\u8fd1\u7dda\u6027\u7684\u6548\u80fd\u64f4\u5c55\u3002<\/p>\n\n\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-full\"><img loading=\"lazy\" decoding=\"async\" width=\"672\" height=\"410\" src=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia03.jpg\" alt=\"\" class=\"wp-image-35926\" title=\"\" srcset=\"https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia03.jpg 672w, https:\/\/www.ioiotimes.com\/wordpress\/wp-content\/uploads\/2023\/06\/20230628-nvidia03-300x183.jpg 300w\" sizes=\"(max-width: 672px) 100vw, 672px\" \/><\/figure><\/div>\n\n\n<p><\/p>\n\n\n\n<p>\u6b64\u5916\uff0cCoreWeave \u5f9e\u96f2\u7aef\u63d0\u4f9b\u7684\u6027\u80fd\u8207 NVIDIA \u5728\u672c\u5730\u8cc7\u6599\u4e2d\u5fc3\u904b\u884c\u7684 AI \u8d85\u7d1a\u96fb\u8166\u5be6\u73fe\u7684\u6027\u80fd\u76f8\u4f3c\u3002\u9019\u8b49\u660e\u4e86 CoreWeave \u6240\u4f7f\u7528\u7684 <a href=\"https:\/\/www.nvidia.com\/en-us\/networking\/quantum2\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA Quantum-2 InfiniBand<\/a> \u7db2\u8def\u5177\u5099\u5730\u4f4e\u5ef6\u9072\u7db2\u8def\u6027\u80fd\u3002<\/p>\n\n\n\n<p>\u5728\u9019\u4e00\u8f2a\u6e2c\u8a66\u4e2d\uff0cMLPerf \u9084\u66f4\u65b0\u4e86\u7528\u65bc\u63a8\u85a6\u7cfb\u7d71\u7684\u57fa\u6e96\u6e2c\u8a66\u3002<\/p>\n\n\n\n<p>\u65b0\u7684\u6e2c\u8a66\u4f7f\u7528\u4e86\u66f4\u5927\u7684\u8cc7\u6599\u96c6\u548c\u66f4\u73fe\u4ee3\u7684 AI \u6a21\u578b\uff0c\u4ee5\u66f4\u597d\u5730\u53cd\u6620\u96f2\u7aef\u670d\u52d9\u4f9b\u61c9\u5546\u9762\u81e8\u7684\u6311\u6230\u3002NVIDIA \u662f\u552f\u4e00\u4e00\u5bb6\u5728\u589e\u5f37\u7684\u57fa\u6e96\u6e2c\u8a66\u4e0a\u63d0\u4ea4\u7d50\u679c\u7684\u516c\u53f8\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u4e0d\u65b7\u64f4\u5c55\u7684<\/strong><strong>NVIDIA AI<\/strong><strong>\u751f\u614b\u7cfb\u7d71<\/strong><\/p>\n\n\n\n<p>\u672c\u8f2a\u6709\u8fd1\u5341\u5e7e\u5bb6\u516c\u53f8\u5728 NVIDIA \u5e73\u53f0\u4e0a\u63d0\u4ea4\u4e86\u7d50\u679c\u3002\u00a0\u4ed6\u5011\u7684\u5de5\u4f5c\u8868\u660e\u4e86\u00a0NVIDIA\u00a0\u4eba\u5de5\u667a\u6167\u5f97\u5230\u4e86\u696d\u754c\u6700\u5ee3\u5927\u7684\u6a5f\u5668\u5b78\u7fd2\u751f\u614b\u7cfb\u7684\u652f\u6301\u3002<\/p>\n\n\n\n<p>\u5305\u62ec Microsoft Azure \u96f2\u7aef\u670d\u52d9\u3001\u7cfb\u7d71\u88fd\u9020\u5546\u83ef\u78a9\u3001\u6234\u723e\u79d1\u6280\u96c6\u5718\u3001\u5bcc\u58eb\u901a\u3001\u6280\u5609\u79d1\u6280\u3001Lenovo\u3001\u96f2\u9054\u79d1\u6280\u548c\u7f8e\u8d85\u5fae\u3002\u8d85\u904e 30 \u500b\u662f\u5728 H100 GPU \u4e0a\u904b\u884c\u7684\u6210\u679c\u3002<\/p>\n\n\n\n<p>\u9019\u7a2e\u7a0b\u5ea6\u7684\u53c3\u8207\u8b93\u4f7f\u7528\u8005\u77e5\u9053\uff0c\u7121\u8ad6\u662f\u5728\u96f2\u7aef\u9084\u662f\u5728\u81ea\u8eab\u8cc7\u6599\u4e2d\u5fc3\u904b\u884c\u7684\u4f3a\u670d\u5668\uff0c\u90fd\u53ef\u4ee5\u900f\u904e NVIDIA\u00a0\u4eba\u5de5\u667a\u6167\u7372\u5f97\u51fa\u8272\u7684\u6548\u80fd\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u8de8\u8d8a\u6240\u6709\u5de5\u4f5c\u8ca0\u8f09\u7684\u6548\u80fd<\/strong><strong><\/strong><\/p>\n\n\n\n<p>NVIDIA&nbsp;\u751f\u614b\u7cfb\u5925\u4f34\u53c3\u8207&nbsp;MLPerf&nbsp;\u662f\u56e0\u70ba\u4ed6\u5011\u77e5\u9053\u9019\u5c0d\u65bc\u5ba2\u6236\u8a55\u4f30\u4eba\u5de5\u667a\u6167\u5e73\u53f0\u548c\u4f9b\u61c9\u5546\u4f86\u8aaa\u662f\u4e00\u500b\u5f88\u6709\u50f9\u503c\u7684\u5de5\u5177\u3002<\/p>\n\n\n\n<p>\u9019\u4e9b\u57fa\u6e96\u6e2c\u8a66\u6db5\u84cb\u4e86\u7528\u6236\u95dc\u5fc3\u7684\u5de5\u4f5c\u8ca0\u8f09\u2014\u2014\u96fb\u8166\u8996\u89ba\u3001\u7ffb\u8b6f\u548c\u5f37\u5316\u5b78\u7fd2\uff0c\u4ee5\u53ca\u751f\u6210\u5f0f\u4eba\u5de5\u667a\u6167\u548c<a href=\"https:\/\/blogs.nvidia.com\/blog\/2020\/05\/14\/whats-a-recommender-system\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u63a8\u85a6\u7cfb\u7d71<\/a>\u3002<\/p>\n\n\n\n<p>\u4f7f\u7528\u8005\u53ef\u4ee5\u4f9d\u9760&nbsp;MLPerf&nbsp;\u7d50\u679c\u505a\u51fa\u660e\u667a\u7684\u8cfc\u8cb7\u6c7a\u5b9a\uff0c\u56e0\u70ba\u6e2c\u8a66\u662f\u900f\u660e\u4e14\u5ba2\u89c0\u7684\u3002&nbsp;\u9019\u4e9b\u57fa\u6e96\u5f97\u5230\u4e86\u5305\u62ec&nbsp;Arm\u3001\u767e\u5ea6\u3001Facebook AI\u3001Google\u3001\u54c8\u4f5b\u5927\u5b78\u3001\u82f1\u7279\u723e\u3001\u5fae\u8edf\u3001\u53f2\u4e39\u798f\u5927\u5b78\u548c\u591a\u502b\u591a\u5927\u5b78\u5728\u5167\u7684\u5ee3\u6cdb\u652f\u6301\u3002<\/p>\n\n\n\n<p>\u73fe\u5728 H100\u3001L4 \u548c <a href=\"https:\/\/www.nvidia.com\/en-us\/autonomous-machines\/embedded-systems\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA Jetson<\/a> \u5e73\u53f0\u4e0a\u505a\u7684 AI \u8a13\u7df4\u3001\u63a8\u8ad6\u548c\u9ad8\u6548\u80fd\u904b\u7b97\u57fa\u6e96\u90fd\u6709 MLPerf \u7684\u6e2c\u8a66\u7d50\u679c\u3002NVIDIA \u4e5f\u5c07\u5728\u672a\u4f86 MLPerf \u4e2d\u63d0\u4ea4\u6e2c\u8a66 <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/grace-hopper-superchip\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA Grace Hopper<\/a> \u7cfb\u7d71\u7684\u7d50\u679c\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><strong>\u80fd\u6e90\u6548\u7387\u7684\u91cd\u8981\u6027<\/strong><strong><\/strong><\/p>\n\n\n\n<p>\u96a8\u8457\u4eba\u5de5\u667a\u6167\u6548\u80fd\u7684\u8981\u6c42\u589e\u9577\uff0c\u63d0\u9ad8\u5be6\u73fe\u6548\u80fd\u7684\u6548\u7387\u81f3\u95dc\u91cd\u8981\u3002&nbsp;\u9019\u5c31\u662f<a href=\"https:\/\/blogs.nvidia.com\/blog\/2021\/09\/01\/what-is-accelerated-computing\/\" target=\"_blank\" rel=\"noreferrer noopener\">\u52a0\u901f\u904b\u7b97<\/a>\u7684\u4f5c\u7528\u3002<\/p>\n\n\n\n<p>\u4f7f\u7528 \u00a0NVIDIA GPU \u52a0\u901f\u7684\u8cc7\u6599\u4e2d\u5fc3\u4f7f\u7528\u66f4\u5c11\u7684\u4f3a\u670d\u5668\u7bc0\u9ede\uff0c\u56e0\u6b64\u4f7f\u7528\u66f4\u5c11\u7684\u6a5f\u67b6\u7a7a\u9593\u548c\u80fd\u6e90\u3002<\/p>\n\n\n\n<p>\u6b64\u5916\uff0c\u52a0\u901f\u7684\u7db2\u8def\u53ef\u4ee5\u63d0\u9ad8\u6548\u7387\u548c\u6548\u80fd\uff0c\u800c\u6301\u7e8c\u7684\u8edf\u9ad4\u512a\u5316\u5247\u80fd\u5728\u76f8\u540c\u7684\u786c\u9ad4\u4e0a\u5be6\u73fe\u4e86\u66f4\u5927\u7684\u6027\u80fd\u63d0\u5347\u3002<\/p>\n\n\n\n<p>\u7bc0\u80fd\u6548\u80fd\u5c0d\u74b0\u5883\u548c\u5546\u696d\u90fd\u975e\u5e38\u6709\u76ca\u3002\u63d0\u5347\u7684\u6548\u53ef\u4ee5\u52a0\u5feb\u4e0a\u5e02\u6642\u9593\uff0c\u8b93\u7d44\u7e54\u80fd\u5920\u5efa\u7acb\u66f4\u5148\u9032\u7684\u61c9\u7528\u7a0b\u5f0f\u3002\u7bc0\u80fd\u6548\u7387\u4e5f\u80fd\u964d\u4f4e\u6210\u672c\uff0c\u56e0\u70ba\u4ee5 NVIDIA GPU \u52a0\u901f\u7684\u8cc7\u6599\u4e2d\u5fc3\u4f7f\u7528\u66f4\u5c11\u7684\u4f3a\u670d\u5668\u7bc0\u9ede\u3002<\/p>\n\n\n\n<p>\u4e8b\u5be6\u4e0a\uff0c\u5728\u6700\u65b0\u00a0<a href=\"https:\/\/www.top500.org\/lists\/green500\/list\/2023\/06\/\" target=\"_blank\" rel=\"noreferrer noopener\">Green500<\/a>\u00a0\u699c\u55ae\u4e0a\u6392\u540d\u524d\u00a030\u00a0\u540d\u7684\u8d85\u7d1a\u96fb\u8166\u4e2d\u6709 22\u00a0\u5957\u63a1\u7528 NVIDIA GPU\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><a><\/a><strong>\u4eba\u4eba\u90fd\u53ef\u4f7f\u7528\u7684\u8edf\u9ad4<\/strong><strong><\/strong><\/p>\n\n\n\n<p>NVIDIA\u00a0\u4eba\u5de5\u667a\u6167\u5e73\u53f0\u7684\u8edf\u9ad4\u5c64 <a href=\"https:\/\/www.nvidia.com\/en-us\/data-center\/products\/ai-enterprise\/\" target=\"_blank\" rel=\"noreferrer noopener\">NVIDIA AI Enterprise<\/a>\u00a0\u53ef\u4ee5\u4f7f\u9818\u5148\u7684\u52a0\u901f\u904b\u7b97\u57fa\u790e\u8a2d\u9054\u81f4\u512a\u5316\u7684\u6548\u80fd\u3002\u00a0\u9019\u500b\u8edf\u9ad4\u914d\u5099\u4e86\u5728\u4f01\u696d\u8cc7\u6599\u4e2d\u5fc3\u904b\u884c\u4eba\u5de5\u667a\u6167\u6240\u9700\u7684\u4f01\u696d\u7d1a\u652f\u6301\u3001\u5b89\u5168\u6027\u548c\u53ef\u9760\u6027\u3002<\/p>\n\n\n\n<p>\u65bc\u6b64\u6b21\u6e2c\u8a66\u4e2d\u4f7f\u7528\u7684\u5404\u7a2e\u8edf\u9ad4\u516c\u958b\u65bc MLPerf \u8cc7\u6e90\u5eab\uff0c\u5e7e\u4e4e\u6bcf\u500b\u4eba\u90fd\u80fd\u53d6\u5f97\u9019\u4e9b\u4e16\u754c\u7d1a\u7684\u6210\u679c\u3002\u6211\u5011\u4e0d\u65b7\u5c07\u6700\u4f73\u5316\u7d50\u679c\u653e\u5165\u00a0<a href=\"https:\/\/ngc.nvidia.com\/catalog\" target=\"_blank\" rel=\"noreferrer noopener\">NGC<\/a>\u00a0\uff08NVIDIA \u7684 GPU\u00a0\u52a0\u901f\u8edf\u9ad4\u76ee\u9304\uff09\u7684\u5bb9\u5668\u4e2d\u3002<\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<p><\/p>\n\n\n\n<h4 class=\"wp-block-heading has-text-align-right has-very-light-gray-to-cyan-bluish-gray-gradient-background has-background\"><br>\ud83d\udfe6<strong>\u73fe\u5728\u5c31\u52a0\u5165&nbsp;<a href=\"https:\/\/www.facebook.com\/profile.php?id=100086628162118\" target=\"_blank\" rel=\"noreferrer noopener\">ioioTIMES \u81c9\u66f8\u7c89\u7d72\u5718<\/a>&nbsp;\u66f4\u591a\u4e92\u52d5\u3001\u66f4\u591a\u597d\u5eb7\u650f\u62b5\u52a0!!<\/strong><br>\ud83d\udfe6<strong>\u6211\u5011\u6709<a href=\"https:\/\/today.line.me\/tw\/v2\/publisher\/103117\" target=\"_blank\" rel=\"noreferrer noopener nofollow\">LINE TODAY<\/a>\u983b\u9053\u4e86\uff0c\u5feb\u4f86\u8ffd\u8e2a\u6211\u5011\u5427!!&#8211;\u6700\u65b0\u79d1\u6280\u65b0\u805e \u76e1\u5728\u4f60\u624b<\/strong><\/h4>\n\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u5728\u5168\u65b0\u696d\u754c\u6a19\u6e96\u57fa\u6e96\u6e2c\u8a66\u4e2d\uff0c\u96f2\u7aef\u670d\u52d9\u4f9b<\/p>\n","protected":false},"author":3,"featured_media":35922,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"rank_math_lock_modified_date":false,"footnotes":""},"categories":[13],"tags":[6361,580,6362,6359,6360,71,5344],"class_list":["post-35906","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-coreweave","tag-focus","tag-gpt-3","tag-h100-gpu","tag-mlperf","tag-nvidia","tag-ai"],"_links":{"self":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/35906"}],"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=35906"}],"version-history":[{"count":5,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/35906\/revisions"}],"predecessor-version":[{"id":35927,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/posts\/35906\/revisions\/35927"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=\/wp\/v2\/media\/35922"}],"wp:attachment":[{"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=35906"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=35906"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.ioiotimes.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=35906"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}