{"id":122992,"date":"2026-11-06T05:56:46","date_gmt":"2026-11-05T21:56:46","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/122992.html"},"modified":"2026-11-06T05:56:46","modified_gmt":"2026-11-05T21:56:46","slug":"gpu-%e5%86%85%e5%ad%98%e6%a8%a1%e5%9e%8b%e6%b5%85%e8%a7%a3-%e5%9b%9b%e5%b1%82%e5%ad%98%e5%82%a8%e3%80%81%e7%bc%96%e7%a8%8b%e6%8e%a7%e5%88%b6","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/122992.html","title":{"rendered":"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236"},"content":{"rendered":"<p>        \/* MD \/ \u5bcc\u6587\u672c\u4e2d\u7684 .toc\uff08\u542b\u535a\u5ba2\u56ed\u642c\u5bb6\u7b49\u5d4c\u5957\u7ed3\u6784\uff09\uff1b.toc-box \u5728\u4fa7\u680f\uff0c\u4e0d\u53d7\u5f71\u54cd *\/<br \/>\n        #content_views .toc,<br \/>\n        \/* \u7f16\u8f91\u5668\u5e38\u5728\u76ee\u5f55\u524d\u540e\u63d2\u5165\u7a7a p\uff08:empty \u4ecd\u5360 20px\uff09\uff0c\u4e00\u5e76\u53bb\u6389\u907f\u514d\u9876\u7a7a\u9699 *\/<br \/>\n        #content_views.markdown_views &gt; p:empty:has(+ .toc),<br \/>\n        #content_views.markdown_views &gt; .toc + p:empty,<br \/>\n        \/* \u5bcc\u6587\u672c\u65e7\u7248\u76ee\u5f55\u6807\u8bb0 *\/<br \/>\n        #content_views.htmledit_views #main-toc,<br \/>\n        #content_views.htmledit_views #hr-toc,<br \/>\n        #content_views.htmledit_views p[id*=\\&#8221;-toc\\&#8221;] {<br \/>\n          display: none !important;<br \/>\n        }<br \/>\n        \/* \u76ee\u5f55\u53bb\u6389\u540e\uff0c\u7d27\u8ddf\u7684\u9996\u4e2a\u6807\u9898\u4e0d\u518d\u591a\u51fa\u4e00\u5757\u4e0a\u8fb9\u8ddd *\/<br \/>\n        #content_views.markdown_views &gt; .toc + h1,<br \/>\n        #content_views.markdown_views &gt; .toc + h2,<br \/>\n        #content_views.markdown_views &gt; .toc + h3,<br \/>\n        #content_views.markdown_views &gt; .toc + h4,<br \/>\n        #content_views.markdown_views &gt; .toc + p:empty + h1,<br \/>\n        #content_views.markdown_views &gt; .toc + p:empty + h2,<br \/>\n        #content_views.markdown_views &gt; .toc + p:empty + h3,<br \/>\n        #content_views.markdown_views &gt; .toc + p:empty + h4 {<br \/>\n          margin-top: 0 !important;<br \/>\n        }<\/p>\n<h2>\u6982\u8ff0<\/h2>\n<p>GPU \u5185\u5b58\u6a21\u578b\u63cf\u8ff0\u7684\u662f&#xff1a;\u6570\u636e\u5b58\u5728\u54ea\u91cc\u3001\u79bb\u8ba1\u7b97\u5355\u5143&#xff08;SM&#xff09;\u6709\u591a\u8fdc\u3001\u8bbf\u95ee\u6709\u591a\u5feb\u3001\u8c01\u6765\u7ba1\u7406\u3002<\/p>\n<p>\u6838\u5fc3\u601d\u60f3\u662f\u5b58\u50a8\u5c42\u6b21\u7ed3\u6784&#xff08;Memory Hierarchy&#xff09;&#xff1a;<\/p>\n<p>\u8d8a\u9760\u8fd1 SM&#xff08;\u8ba1\u7b97\u6838\u5fc3&#xff09; \u2192 \u8d8a\u5feb\u3001\u8d8a\u5c0f\u3001\u8d8a\u8d35<br \/>\n\u8d8a\u8fdc\u79bb SM           \u2192 \u8d8a\u6162\u3001\u8d8a\u5927\u3001\u8d8a\u4fbf\u5b9c<\/p>\n<p>A100 \u7684\u56db\u5c42\u7ed3\u6784&#xff1a;<br \/>\n<img decoding=\"async\" src=\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/11\/20261105215645-6aecfb9d3cd63.png\" alt=\"\u5728\u8fd9\u91cc\u63d2\u5165\u56fe\u7247\u63cf\u8ff0\" \/><\/p>\n<p>\u5c0f\u7ed3: \u5bc4\u5b58\u5668 1 \u5468\u671f&#xff0c;\u5171\u4eab\u5185\u5b58 33 \u5468\u671f&#xff0c;L2 200 \u5468\u671f&#xff0c;\u5168\u5c40\u5185\u5b58 290 \u5468\u671f\u3002\u4ece\u5171\u4eab\u5185\u5b58\u5230\u5168\u5c40\u5185\u5b58&#xff0c;\u5ef6\u8fdf\u5dee\u4e86\u8fd1 9 \u500d\u3002<\/p>\n<p>\u5982\u679c\u8ba1\u7b97\u5355\u5143\u7b49\u5168\u5c40\u5185\u5b58&#xff0c;\u5927\u91cf\u65f6\u95f4\u5728\u7a7a\u8f6c\u3002<\/p>\n<h2>\u56db\u5c42\u5b58\u50a8\u7684\u4f5c\u7528\u4e0e\u7f16\u7a0b\u63a7\u5236<\/h2>\n<h3>\u5168\u5c40\u5185\u5b58&#xff08;Global Memory&#xff09;<\/h3>\n<table>\n<tr>\u7279\u6027\u53c2\u6570<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u4f4d\u7f6e<\/td>\n<td align=\"left\">GPU \u82af\u7247\u5916 HBM \u663e\u5b58<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5bb9\u91cf<\/td>\n<td align=\"left\">A100 80GB<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5e26\u5bbd<\/td>\n<td align=\"left\">\u7ea6 2 TB\/s<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5ef6\u8fdf<\/td>\n<td align=\"left\">\u7ea6 290 \u5468\u671f<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7f16\u7a0b\u63a7\u5236<\/td>\n<td align=\"left\">\u624b\u52a8 cudaMalloc&#xff0c;\u6240\u6709\u7ebf\u7a0b\u53ef\u89c1<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u4f5c\u7528&#xff1a; \u5b58\u653e\u6a21\u578b\u6743\u91cd\u3001\u6fc0\u6d3b\u503c\u3001\u68af\u5ea6\u3001\u8bad\u7ec3\u6570\u636e\u3001\u4e2d\u95f4\u7ed3\u679c\u3001\u6700\u7ec8\u8f93\u51fa\u3002\u662f GPU \u7684\u201c\u5927\u4ed3\u5e93\u201d&#xff0c;\u5bb9\u91cf\u5927\u4f46\u901f\u5ea6\u6162\u3002<\/p>\n<p>\u7f16\u7a0b\u63a7\u5236&#xff1a; \u9700\u8981\u624b\u52a8\u5206\u914d\u3001\u624b\u52a8\u62f7\u8d1d&#xff08;CPU\u2192GPU \u901a\u8fc7 PCIe&#xff09;\u3002\u6240\u6709\u7ebf\u7a0b\u90fd\u80fd\u8bbf\u95ee&#xff0c;\u4f46\u6bcf\u6b21\u8bbf\u95ee\u90fd\u8981\u4ed8\u51fa\u9ad8\u5ef6\u8fdf\u3002<\/p>\n<h3>L2 \u7f13\u5b58<\/h3>\n<table>\n<tr>\u7279\u6027\u53c2\u6570<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u4f4d\u7f6e<\/td>\n<td align=\"left\">GPU \u82af\u7247\u5185&#xff0c;\u6240\u6709 SM \u5171\u4eab<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5bb9\u91cf<\/td>\n<td align=\"left\">40MB (A100)<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5e26\u5bbd<\/td>\n<td align=\"left\">\u7ea6 5 TB\/s<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5ef6\u8fdf<\/td>\n<td align=\"left\">\u7ea6 200 \u5468\u671f<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7f16\u7a0b\u63a7\u5236<\/td>\n<td align=\"left\">\u786c\u4ef6\u81ea\u52a8\u7ba1\u7406&#xff0c;\u4e0d\u53ef\u624b\u52a8\u63a7\u5236<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u4f5c\u7528&#xff1a;<\/p>\n<ul>\n<li>\u7f13\u5b58\u5168\u5c40\u5185\u5b58\u7684\u70ed\u70b9\u6570\u636e&#xff08;\u5982\u9891\u7e41\u8bbf\u95ee\u7684\u6743\u91cd&#xff09;&#xff0c;\u907f\u514d\u6bcf\u6b21\u90fd\u8bbf\u95ee HBM&#xff1b;<\/li>\n<li>\u6240\u6709 SM \u4e0e HBM \u4e4b\u95f4\u7684\u6570\u636e\u901a\u4fe1\u90fd\u5fc5\u987b\u7ecf\u8fc7 L2&#xff0c;\u662f GPU \u7684\u201c\u6570\u636e\u603b\u67a2\u7ebd\u201d&#xff1b;<\/li>\n<li>\u4e0d\u540c SM \u4e0a\u7684\u7ebf\u7a0b\u53ef\u4ee5\u901a\u8fc7 L2 \u9ad8\u6548\u5171\u4eab\u6570\u636e&#xff0c;\u4fdd\u8bc1\u6570\u636e\u4e00\u81f4\u6027\u3002<\/li>\n<\/ul>\n<p>\u7f16\u7a0b\u63a7\u5236&#xff1a; \u5b8c\u5168\u81ea\u52a8&#xff0c;\u7a0b\u5e8f\u5458\u65e0\u6cd5\u663e\u5f0f\u6307\u5b9a\u201c\u628a\u67d0\u6570\u636e\u653e L2\u201d\u3002\u4f46\u53ef\u4ee5\u901a\u8fc7\u8bbf\u95ee\u6a21\u5f0f\u5f71\u54cd L2 \u547d\u4e2d\u7387\u3002<\/p>\n<h3>L1 \u7f13\u5b58 \/ \u5171\u4eab\u5185\u5b58<\/h3>\n<table>\n<tr>\u7279\u6027\u53c2\u6570<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u4f4d\u7f6e<\/td>\n<td align=\"left\">\u6bcf\u4e2a SM \u5185\u90e8<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5bb9\u91cf<\/td>\n<td align=\"left\">192KB\/SM<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5ef6\u8fdf<\/td>\n<td align=\"left\">\u7ea6 33 \u5468\u671f<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7f16\u7a0b\u63a7\u5236<\/td>\n<td align=\"left\">\u5171\u4eab\u5185\u5b58\u624b\u52a8 __shared__&#xff1b;L1 \u81ea\u52a8<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u4f5c\u7528&#xff1a;<\/p>\n<ul>\n<li>\u5171\u4eab\u5185\u5b58&#xff1a; Block \u5185\u7ebf\u7a0b\u534f\u4f5c\u7684\u201c\u5c0f\u4ed3\u5e93\u201d&#xff0c;\u7528\u4e8e\u4ea4\u6362\u6570\u636e&#xff08;\u5982\u77e9\u9635\u4e58\u6cd5\u7684\u5206\u5757\u6570\u636e Tiling&#xff09;&#xff1b;<\/li>\n<li>L1 \u7f13\u5b58&#xff1a; \u4e0d\u624b\u52a8\u7528\u5171\u4eab\u5185\u5b58\u65f6&#xff0c;\u81ea\u52a8\u7f13\u5b58\u5168\u5c40\u5185\u5b58\u6570\u636e\u3002<\/li>\n<\/ul>\n<p>\u7f16\u7a0b\u63a7\u5236&#xff1a; \u5171\u4eab\u5185\u5b58\u5b8c\u5168\u624b\u52a8\u3002\u7a0b\u5e8f\u5458\u7528 __shared__ \u58f0\u660e&#xff0c;\u7528 __syncthreads() \u540c\u6b65\u3002\u8fd9\u662f GPU \u6027\u80fd\u4f18\u5316\u7684\u6838\u5fc3\u624b\u6bb5\u3002<\/p>\n<h3>\u5bc4\u5b58\u5668\u6587\u4ef6<\/h3>\n<table>\n<tr>\u7279\u6027\u53c2\u6570<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u4f4d\u7f6e<\/td>\n<td align=\"left\">SM \u5185&#xff0c;\u6bcf\u4e2a CUDA \u6838\u5fc3\u65c1<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5bb9\u91cf<\/td>\n<td align=\"left\">256KB\/SM<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5e26\u5bbd<\/td>\n<td align=\"left\">\u7ea6 10 TB\/s&#xff08;\u7406\u8bba&#xff09;<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5ef6\u8fdf<\/td>\n<td align=\"left\">1 \u5468\u671f&#xff08;\u96f6\u5f00\u9500&#xff09;<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7f16\u7a0b\u63a7\u5236<\/td>\n<td align=\"left\">\u7f16\u8bd1\u5668\u81ea\u52a8\u5206\u914d&#xff0c;\u7ebf\u7a0b\u79c1\u6709<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u4f5c\u7528&#xff1a; \u5b58\u50a8\u7ebf\u7a0b\u7684\u5c40\u90e8\u53d8\u91cf\u3001\u4e34\u65f6\u7ed3\u679c\u3002\u6bcf\u4e2a\u7ebf\u7a0b\u6700\u591a 255 \u4e2a\u5bc4\u5b58\u5668\u3002\u901f\u5ea6\u6781\u5feb&#xff0c;\u4f46\u5bb9\u91cf\u6709\u9650\u3002<\/p>\n<p>\u7f16\u7a0b\u63a7\u5236&#xff1a; \u5b8c\u5168\u81ea\u52a8&#xff0c;\u7f16\u8bd1\u5668\u51b3\u5b9a\u53d8\u91cf\u653e\u5bc4\u5b58\u5668\u8fd8\u662f\u6ea2\u51fa\u5230\u672c\u5730\u5185\u5b58\u3002\u7a0b\u5e8f\u5458\u53ef\u4ee5\u901a\u8fc7\u51cf\u5c11\u53d8\u91cf\u3001\u7b80\u5316\u4ee3\u7801\u6765\u964d\u4f4e\u5bc4\u5b58\u5668\u7528\u91cf&#xff0c;\u4ece\u800c\u8ba9 SM \u9a7b\u7559\u66f4\u591a Warp\u3002<\/p>\n<h2>\u77e9\u9635\u4e58\u6cd5\u5206\u5757&#xff0c;\u5c55\u793a\u56db\u5c42\u5b58\u50a8\u7684\u5a01\u529b<\/h2>\n<h3>\u95ee\u9898\u8bbe\u5b9a<\/h3>\n<p>\u8ba1\u7b97 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">C&#061;A\u00d7BC &#061; A \\\\times B<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7667em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\">A<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0502em\">B<\/span><\/span><\/span><\/span><\/span>&#xff0c;\u5176\u4e2d <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">AA<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\">A<\/span><\/span><\/span><\/span><\/span>\u3001<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">BB<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0502em\">B<\/span><\/span><\/span><\/span><\/span> \u662f <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">N\u00d7NN \\\\times N<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7667em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><\/span> \u77e9\u9635&#xff0c;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">N&#061;1024N &#061; 1024<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">1024<\/span><\/span><\/span><\/span><\/span>\u3002<\/p>\n<p>\u6bcf\u4e2a\u8f93\u51fa\u5143\u7d20 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">C[i][j]C[i][j]<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">]<\/span><\/span><\/span><\/span><\/span> \u9700\u8981&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">C[i][j]&#061;\u2211k&#061;0N\u22121A[i][k]\u00d7B[k][j]<br \/>\nC[i][j] &#061; \\\\sum_{k&#061;0}^{N-1} A[i][k] \\\\times B[k][j]<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">]<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.1304em;vertical-align: -1.3021em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8283em\"><span class=\"\" style=\"top: -1.8479em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.109em\">N<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3021em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">A<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">]<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0502em\">B<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">]<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5171 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">NN<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><\/span> \u6b21\u4e58\u52a0\u3002\u603b\u8ba1\u7b97\u91cf&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">N3&#061;10243\u22481.07\u00d7109\u00a0\u6b21\u4e58\u52a0<br \/>\nN^3 &#061; 1024^3 \\\\approx 1.07 \\\\times 10^9 \\\\text{ \u6b21\u4e58\u52a0}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">3<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">102<\/span><span class=\"mord\"><span class=\"mord\">4<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">3<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1.07<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u6b21\u4e58\u52a0<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<h3>\u65b9\u6848\u4e00&#xff1a;\u4e0d\u5206\u5757&#xff0c;\u76f4\u63a5\u4ece\u5168\u5c40\u5185\u5b58\u8bfb<\/h3>\n<p>\u6bcf\u4e2a\u7ebf\u7a0b\u8ba1\u7b97\u4e00\u4e2a <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">C[i][j]C[i][j]<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">]<\/span><\/span><\/span><\/span><\/span>&#xff0c;\u5faa\u73af <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">kk<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><\/span><\/span><\/span><\/span> \u4ece 0 \u5230 1023&#xff1a;<\/p>\n<p>for (int k &#061; 0; k &lt; N; k&#043;&#043;) {<br \/>\n    C[i][j] &#043;&#061; A[i][k] * B[k][j];   \/\/ \u6bcf\u6b21\u4ece\u5168\u5c40\u5185\u5b58\u8bfb A \u548c B<br \/>\n}<\/p>\n<p>\u5168\u5c40\u5185\u5b58\u8bbf\u95ee\u6b21\u6570&#xff1a;<\/p>\n<ul>\n<li>\u6bcf\u4e2a\u7ebf\u7a0b\u8bfb <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">A[i][k]A[i][k]<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">A<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">]<\/span><\/span><\/span><\/span><\/span> \u5171 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">NN<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><\/span> \u6b21&#xff0c;\u8bfb <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">B[k][j]B[k][j]<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0502em\">B<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">]<\/span><span class=\"mopen\">[<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">]<\/span><\/span><\/span><\/span><\/span> \u5171 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">NN<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><\/span> \u6b21&#xff1b;<\/li>\n<li>\u603b\u7ebf\u7a0b\u6570 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">N2&#061;10242\u22481.05\u00d7106N^2 &#061; 1024^2 \\\\approx 1.05 \\\\times 10^6<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8141em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8141em\"><\/span><span class=\"mord\">102<\/span><span class=\"mord\"><span class=\"mord\">4<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1.05<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8141em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">6<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1b;<\/li>\n<li>\u603b\u8bfb\u53d6\u6b21\u6570&#xff1a;<\/li>\n<\/ul>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">2\u00d7N3&#061;2\u00d71.07\u00d7109\u22482.14\u00d7109\u00a0\u6b21<br \/>\n2 \\\\times N^3 &#061; 2 \\\\times 1.07 \\\\times 10^9 \\\\approx 2.14 \\\\times 10^9 \\\\text{ \u6b21}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">3<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1.07<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2.14<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u6b21<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u6bcf\u6b21\u8bbf\u95ee\u5ef6\u8fdf\u7ea6 290 \u5468\u671f\u3002\u5373\u4f7f\u6709 L2 \u7f13\u5b58&#xff0c;\u5927\u91cf\u91cd\u590d\u8bfb\u53d6\u4e5f\u4f1a\u9020\u6210\u4e25\u91cd\u74f6\u9888\u3002<\/p>\n<p>\u4f30\u7b97\u65f6\u95f4&#xff08;\u4e0d\u8003\u8651\u7f13\u5b58\u547d\u4e2d&#xff09;&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">2.14\u00d7109\u00d7290\u00a0\u5468\u671f\u22486.2\u00d71011\u00a0\u5468\u671f<br \/>\n2.14 \\\\times 10^9 \\\\times 290 \\\\text{ \u5468\u671f} \\\\approx 6.2 \\\\times 10^{11} \\\\text{ \u5468\u671f}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2.14<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.9474em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord\">290<\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u5468\u671f<\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">6.2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">11<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u5468\u671f<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>A100 \u7ea6 1.4 GHz&#xff0c;\u65f6\u95f4\u7ea6&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">6.2\u00d71011\/1.4\u00d7109\u2248443\u00a0\u79d2<br \/>\n6.2 \\\\times 10^{11} \/ 1.4 \\\\times 10^9 \\\\approx 443 \\\\text{ \u79d2}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">6.2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1141em;vertical-align: -0.25em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">11<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord\">\/1.4<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord\">443<\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u79d2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u8fd9\u8fd8\u53ea\u662f\u5185\u5b58\u8bbf\u95ee&#xff0c;\u8ba1\u7b97\u672c\u8eab\u8fdc\u5c0f\u4e8e\u8fd9\u4e2a\u6570\u3002<\/p>\n<h3>\u65b9\u6848\u4e8c&#xff1a;\u5206\u5757\u52a0\u8f7d\u5230\u5171\u4eab\u5185\u5b58&#xff08;Tiling&#xff09;<\/h3>\n<p>\u628a\u77e9\u9635\u5207\u6210 <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">T\u00d7TT \\\\times T<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7667em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">T<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><\/span> \u7684\u5c0f\u5757&#xff0c;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">T&#061;32T &#061; 32<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">T<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span>\u3002\u6bcf\u4e2a Block \u8d1f\u8d23\u8ba1\u7b97 C \u7684\u4e00\u4e2a <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">32\u00d73232 \\\\times 32<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span> \u5757\u3002<\/p>\n<p>\u6bcf\u4e2a Block \u7684\u7ebf\u7a0b\u534f\u4f5c&#xff1a;<\/p>\n<li>\u628a A \u7684\u4e00\u4e2a <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">32\u00d73232 \\\\times 32<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span> \u5757\u548c B \u7684\u4e00\u4e2a <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">32\u00d73232 \\\\times 32<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span> \u5757\u52a0\u8f7d\u5230\u5171\u4eab\u5185\u5b58&#xff1b;<\/li>\n<li>\u7528\u5171\u4eab\u5185\u5b58\u505a\u4e58\u52a0&#xff1b;<\/li>\n<li>\u6cbf k \u65b9\u5411\u6ed1\u52a8&#xff0c;\u91cd\u590d <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">N\/T&#061;32N\/T &#061; 32<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"mord\">\/<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">T<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span> \u6b21\u3002<\/li>\n<p>__shared__ float As[32][32];<br \/>\n__shared__ float Bs[32][32];<\/p>\n<p>for (int t &#061; 0; t &lt; N\/32; t&#043;&#043;) {<br \/>\n    \/\/ \u534f\u4f5c\u52a0\u8f7d A \u548c B \u7684\u5757\u5230\u5171\u4eab\u5185\u5b58<br \/>\n    As[ty][tx] &#061; A[i][t*32 &#043; tx];<br \/>\n    Bs[ty][tx] &#061; B[t*32 &#043; ty][j];<br \/>\n    __syncthreads();<\/p>\n<p>    \/\/ \u7528\u5171\u4eab\u5185\u5b58\u505a 32 \u6b21\u4e58\u52a0<br \/>\n    for (int k &#061; 0; k &lt; 32; k&#043;&#043;) {<br \/>\n        C[i][j] &#043;&#061; As[ty][k] * Bs[k][tx];<br \/>\n    }<br \/>\n    __syncthreads();<br \/>\n}<\/p>\n<p>\u5168\u5c40\u5185\u5b58\u8bbf\u95ee\u6b21\u6570&#xff1a;<\/p>\n<ul>\n<li>\u6bcf\u4e2a Block \u52a0\u8f7d A \u5757\u548c B \u5757&#xff0c;\u6bcf\u6b21 32\u00d732 \u4e2a\u5143\u7d20&#xff1b;<\/li>\n<li>\u5171 N\/T &#061; 32 \u6b21\u6ed1\u52a8&#xff1b;<\/li>\n<li>\u6bcf\u4e2a Block \u5168\u5c40\u8bfb\u53d6&#xff1a;<\/li>\n<\/ul>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">2\u00d732\u00d732\u00d732&#061;2\u00d732768&#061;65536\u00a0\u6b21<br \/>\n2 \\\\times 32 \\\\times 32 \\\\times 32 &#061; 2 \\\\times 32768 &#061; 65536 \\\\text{ \u6b21}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32768<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord\">65536<\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u6b21<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<ul>\n<li>Block \u603b\u6570&#xff1a;<\/li>\n<\/ul>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">(N\/32)2&#061;322&#061;1024<br \/>\n(N\/32)^2 &#061; 32^2 &#061; 1024<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.1141em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"mord\">\/32<\/span><span class=\"mclose\"><span class=\"mclose\">)<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">3<\/span><span class=\"mord\"><span class=\"mord\">2<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">1024<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<ul>\n<li>\u603b\u5168\u5c40\u8bfb\u53d6&#xff1a;<\/li>\n<\/ul>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">1024\u00d765536\u22486.7\u00d7107\u00a0\u6b21<br \/>\n1024 \\\\times 65536 \\\\approx 6.7 \\\\times 10^7 \\\\text{ \u6b21}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1024<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">65536<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">6.7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">7<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u6b21<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5bf9\u6bd4\u65b9\u6848\u4e00&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">2.14\u00d71096.7\u00d7107\u224832<br \/>\n\\\\frac{2.14 \\\\times 10^9}{6.7 \\\\times 10^7} \\\\approx 32<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 2.2604em;vertical-align: -0.7693em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.4911em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">6.7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7401em\"><span class=\"\" style=\"top: -2.989em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">7<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">2.14<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7693em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5168\u5c40\u5185\u5b58\u8bbf\u95ee\u51cf\u5c11\u4e86 32 \u500d\u3002<\/p>\n<p>\u5171\u4eab\u5185\u5b58\u8bbf\u95ee\u6b21\u6570&#xff1a;<\/p>\n<p>\u6bcf\u4e2a\u7ebf\u7a0b\u505a <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">NN<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><\/span> \u6b21\u4e58\u52a0&#xff0c;\u6bcf\u6b21\u8bfb\u5171\u4eab\u5185\u5b58&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">N3&#061;1.07\u00d7109\u00a0\u6b21\u5171\u4eab\u5185\u5b58\u8bfb\u53d6<br \/>\nN^3 &#061; 1.07 \\\\times 10^9 \\\\text{ \u6b21\u5171\u4eab\u5185\u5b58\u8bfb\u53d6}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.109em\">N<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">3<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1.07<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u6b21\u5171\u4eab\u5185\u5b58\u8bfb\u53d6<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5171\u4eab\u5185\u5b58\u5ef6\u8fdf 33 \u5468\u671f&#xff0c;\u8fdc\u4f4e\u4e8e\u5168\u5c40\u5185\u5b58 290 \u5468\u671f\u3002<\/p>\n<p>\u4f30\u7b97\u65f6\u95f4&#xff1a;<\/p>\n<ul>\n<li>\u5168\u5c40\u5185\u5b58&#xff1a;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">6.7\u00d7107\u00d7290\u22481.94\u00d710106.7 \\\\times 10^7 \\\\times 290 \\\\approx 1.94 \\\\times 10^{10}<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">6.7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8974em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">7<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">290<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1.94<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8141em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">10<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> \u5468\u671f&#xff1b;<\/li>\n<li>\u5171\u4eab\u5185\u5b58&#xff1a;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">1.07\u00d7109\u00d733\u22483.53\u00d710101.07 \\\\times 10^9 \\\\times 33 \\\\approx 3.53 \\\\times 10^{10}<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1.07<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8974em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">33<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">3.53<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8141em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8141em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">10<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> \u5468\u671f&#xff1b;<\/li>\n<li>\u4f46\u5171\u4eab\u5185\u5b58\u8bbf\u95ee\u53ef\u4ee5\u6d41\u6c34\u7ebf\u5316&#xff0c;\u4e14 Warp \u5207\u6362\u80fd\u9690\u85cf\u5ef6\u8fdf\u3002<\/li>\n<\/ul>\n<p>\u603b\u65f6\u95f4\u7ea6&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">(1.94&#043;3.53)\u00d71010\/1.4\u00d7109\u224839\u00a0\u79d2<br \/>\n(1.94 &#043; 3.53) \\\\times 10^{10} \/ 1.4 \\\\times 10^9 \\\\approx 39 \\\\text{ \u79d2}<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">1.94<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\">3.53<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1141em;vertical-align: -0.25em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">10<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord\">\/1.4<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8641em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">9<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord\">39<\/span><span class=\"mord text\"><span class=\"mord\">\u00a0<\/span><span class=\"mord cjk_fallback\">\u79d2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5bf9\u6bd4\u65b9\u6848\u4e00 443 \u79d2&#xff0c;\u52a0\u901f\u7ea6 11 \u500d\u3002\u8fd9\u5c31\u662f\u5171\u4eab\u5185\u5b58\u5206\u5757\u7684\u4ef7\u503c\u3002<\/p>\n<h3>\u65b9\u6848\u4e09&#xff1a;\u5bc4\u5b58\u5668\u8fdb\u4e00\u6b65\u4f18\u5316<\/h3>\n<p>\u6bcf\u4e2a\u7ebf\u7a0b\u8ba1\u7b97 C \u7684\u4e00\u4e2a 4\u00d74 \u5c0f\u5757&#xff08;\u800c\u4e0d\u662f\u5355\u4e2a\u5143\u7d20&#xff09;&#xff0c;\u628a A\u3001B \u7684\u4e2d\u95f4\u7ed3\u679c\u653e\u5728\u5bc4\u5b58\u5668&#xff1a;<\/p>\n<p>float c[4][4] &#061; {0};   \/\/ 16 \u4e2a\u5bc4\u5b58\u5668<br \/>\nfloat a[4], b[4];       \/\/ 8 \u4e2a\u5bc4\u5b58\u5668<\/p>\n<p>\u6bcf\u4e2a\u7ebf\u7a0b\u7528 24 \u4e2a\u5bc4\u5b58\u5668&#xff0c;\u5171\u4eab\u5185\u5b58\u8bfb\u53d6\u6b21\u6570\u51cf\u5c11&#xff1a;<\/p>\n<ul>\n<li>\u539f\u6765\u6bcf\u4e2a\u7ebf\u7a0b\u8bfb 2N \u6b21\u5171\u4eab\u5185\u5b58&#xff1b;<\/li>\n<li>\u73b0\u5728\u6bcf\u4e2a\u7ebf\u7a0b\u8ba1\u7b97 16 \u4e2a\u8f93\u51fa&#xff0c;\u5171\u4eab\u5185\u5b58\u8bfb\u53d6\u5206\u644a\u5230 16 \u4e2a\u8f93\u51fa\u4e0a\u3002<\/li>\n<\/ul>\n<p>\u5171\u4eab\u5185\u5b58\u8bbf\u95ee\u6b21\u6570\u4ece N^3 \u964d\u5230\u7ea6 N^3\/4&#xff0c;\u8fdb\u4e00\u6b65\u51cf\u5c11\u3002<\/p>\n<p>\u5bc4\u5b58\u5668\u5ef6\u8fdf 1 \u5468\u671f&#xff0c;\u51e0\u4e4e\u96f6\u5f00\u9500\u3002\u4f46\u5bc4\u5b58\u5668\u7528\u91cf\u589e\u52a0\u4f1a\u9650\u5236 SM \u9a7b\u7559\u7684 Warp \u6570&#xff1a;<\/p>\n<ul>\n<li>\u6bcf\u4e2a\u7ebf\u7a0b 24 \u4e2a\u5bc4\u5b58\u5668&#xff0c;SM \u6709 65536 \u4e2a\u5bc4\u5b58\u5668&#xff0c;\u6700\u591a\u9a7b\u7559 65536\/(24 \u00d7 32) \u2248 85 \u4e2a Warp&#xff1b;<\/li>\n<li>\u5b9e\u9645\u53d7 64 \u4e2a Warp \u69fd\u4f4d\u9650\u5236&#xff0c;\u4ecd\u80fd\u6ee1\u8d1f\u8377\u3002<\/li>\n<\/ul>\n<p>\u5982\u679c\u6bcf\u4e2a\u7ebf\u7a0b\u7528 128 \u4e2a\u5bc4\u5b58\u5668&#xff0c;\u6700\u591a\u9a7b\u7559 65536\/(128 \u00d7 32) &#061; 16 \u4e2a Warp&#xff0c;\u5e76\u884c\u5ea6\u4e0b\u964d&#xff0c;\u5ef6\u8fdf\u9690\u85cf\u80fd\u529b\u53d8\u5dee\u3002<\/p>\n<h2>\u4e3a\u4ec0\u4e48 GPU \u5185\u5b58\u8981\u5206\u8fd9\u4e48\u591a\u5c42&#xff1f;<\/h2>\n<h3>\u7269\u7406\u539f\u56e0&#xff1a;\u8ddd\u79bb\u51b3\u5b9a\u5ef6\u8fdf<\/h3>\n<p>\u7535\u4fe1\u53f7\u5728\u5bfc\u7ebf\u4e2d\u4f20\u64ad\u9700\u8981\u65f6\u95f4&#xff0c;\u7269\u7406\u8ddd\u79bb\u8d8a\u77ed&#xff0c;\u5ef6\u8fdf\u8d8a\u4f4e&#xff1a;<\/p>\n<ul>\n<li>\u5bc4\u5b58\u5668\u5728 CUDA \u6838\u5fc3\u65c1&#xff0c;\u8ddd\u79bb\u6700\u8fd1&#xff0c;1 \u5468\u671f&#xff1b;<\/li>\n<li>\u5171\u4eab\u5185\u5b58\u5728 SM \u5185&#xff0c;33 \u5468\u671f&#xff1b;<\/li>\n<li>L2 \u5728\u82af\u7247\u5185\u4f46\u8de8 SM&#xff0c;200 \u5468\u671f&#xff1b;<\/li>\n<li>\u5168\u5c40\u5185\u5b58\u5728\u82af\u7247\u5916 HBM&#xff0c;290 \u5468\u671f\u3002<\/li>\n<\/ul>\n<h3>\u5bb9\u91cf\u4e0e\u6210\u672c\u7684\u6743\u8861<\/h3>\n<table>\n<tr>\u5c42\u7ea7\u5bb9\u91cf\u6210\u672c\u901f\u5ea6<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u5bc4\u5b58\u5668<\/td>\n<td align=\"left\">256KB\/SM<\/td>\n<td align=\"left\">\u6781\u8d35<\/td>\n<td align=\"left\">1 \u5468\u671f<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5171\u4eab\u5185\u5b58<\/td>\n<td align=\"left\">192KB\/SM<\/td>\n<td align=\"left\">\u8d35<\/td>\n<td align=\"left\">33 \u5468\u671f<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">L2<\/td>\n<td align=\"left\">40MB<\/td>\n<td align=\"left\">\u4e2d<\/td>\n<td align=\"left\">200 \u5468\u671f<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5168\u5c40\u5185\u5b58<\/td>\n<td align=\"left\">80GB<\/td>\n<td align=\"left\">\u76f8\u5bf9\u4fbf\u5b9c<\/td>\n<td align=\"left\">290 \u5468\u671f<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u82af\u7247\u9762\u79ef\u6709\u9650&#xff0c;\u4e0d\u53ef\u80fd\u628a 80GB \u90fd\u505a\u6210\u5bc4\u5b58\u5668\u3002\u53ea\u80fd\u5206\u5c42&#xff1a;\u5c0f\u5bb9\u91cf\u9ad8\u901f\u5b58\u50a8\u9760\u8fd1\u8ba1\u7b97&#xff0c;\u5927\u5bb9\u91cf\u4f4e\u901f\u5b58\u50a8\u653e\u5916\u9762\u3002<\/p>\n<h3>\u5c40\u90e8\u6027\u539f\u7406<\/h3>\n<ul>\n<li>\u65f6\u95f4\u5c40\u90e8\u6027&#xff1a;\u521a\u8bbf\u95ee\u7684\u6570\u636e\u5f88\u53ef\u80fd\u518d\u6b21\u8bbf\u95ee&#xff08;\u5982\u5faa\u73af\u4e2d\u7684\u6743\u91cd&#xff09;\u2192 L2 \u7f13\u5b58&#xff1b;<\/li>\n<li>\u7a7a\u95f4\u5c40\u90e8\u6027&#xff1a;\u76f8\u90bb\u6570\u636e\u5f88\u53ef\u80fd\u4e00\u8d77\u8bbf\u95ee&#xff08;\u5982\u77e9\u9635\u540c\u884c\u5143\u7d20&#xff09;\u2192 \u5171\u4eab\u5185\u5b58\u5206\u5757&#xff1b;<\/li>\n<li>\u5e7f\u64ad\u7279\u6027&#xff1a;Warp \u5185 32 \u7ebf\u7a0b\u8bfb\u540c\u4e00\u5e38\u91cf \u2192 \u5e38\u91cf\u5185\u5b58\u4e00\u6b21\u8bfb\u53d6\u670d\u52a1 32 \u7ebf\u7a0b\u3002<\/li>\n<\/ul>\n<h3>\u6570\u503c\u76f4\u89c9&#xff1a;\u5ef6\u8fdf\u9690\u85cf<\/h3>\n<p>\u8bbe\u4e00\u4e2a Warp \u8bbf\u95ee\u5168\u5c40\u5185\u5b58\u9700\u8981 290 \u5468\u671f&#xff0c;\u8ba1\u7b97\u9700\u8981 4 \u5468\u671f\u3002<\/p>\n<p>\u5982\u679c SM \u53ea\u9a7b\u7559 1 \u4e2a Warp&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">\u5229\u7528\u7387&#061;44&#043;290\u22481.4%\u5229\u7528\u7387 &#061; \\\\frac{4}{4 &#043; 290} \\\\approx 1.4\\\\%<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord cjk_fallback\">\u5229\u7528\u7387<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0908em;vertical-align: -0.7693em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3214em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">4<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">290<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">4<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7693em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8056em;vertical-align: -0.0556em\"><\/span><span class=\"mord\">1.4%<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5982\u679c\u9a7b\u7559 64 \u4e2a Warp&#xff0c;\u8f6e\u6d41\u4e0a\u8ba1\u7b97\u5355\u5143&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">\u5229\u7528\u7387\u2248100%<br \/>\n\u5229\u7528\u7387 \\\\approx 100\\\\%<br \/>\n<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord cjk_fallback\">\u5229\u7528\u7387<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8056em;vertical-align: -0.0556em\"><\/span><span class=\"mord\">100%<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u6240\u4ee5 GPU \u7528\u201c\u5927\u91cf Warp \u5e76\u884c\u201d\u6765\u9690\u85cf\u9ad8\u5ef6\u8fdf&#xff0c;\u8fd9\u662f\u541e\u5410\u91cf\u4f18\u5316\u7684\u6838\u5fc3\u3002<\/p>\n<h2>GPU \u5185\u5b58\u4e0e CPU \u5185\u5b58\u7684\u672c\u8d28\u533a\u522b<\/h2>\n<table>\n<tr>\u7279\u6027GPU (A100)CPU (Xeon)<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u4e3b\u5b58\u5e26\u5bbd<\/td>\n<td align=\"left\">\u7ea6 2 TB\/s<\/td>\n<td align=\"left\">\u6570\u767e GB\/s<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7f13\u5b58\u63a7\u5236<\/td>\n<td align=\"left\">\u5171\u4eab\u5185\u5b58\u624b\u52a8\u63a7\u5236<\/td>\n<td align=\"left\">\u7f13\u5b58\u5b8c\u5168\u81ea\u52a8<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u7ebf\u7a0b\u5bc4\u5b58\u5668<\/td>\n<td align=\"left\">255 \u4e2a\/\u7ebf\u7a0b<\/td>\n<td align=\"left\">16 \u4e2a\/\u7ebf\u7a0b&#xff08;x86-64&#xff09;<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5ef6\u8fdf\u5bb9\u5fcd<\/td>\n<td align=\"left\">\u901a\u8fc7 Warp \u5207\u6362\u9690\u85cf\u5ef6\u8fdf<\/td>\n<td align=\"left\">\u5c3d\u91cf\u964d\u4f4e\u5ef6\u8fdf<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5185\u5b58\u6a21\u578b<\/td>\n<td align=\"left\">\u5171\u4eab\u5185\u5b58\u663e\u5f0f\u540c\u6b65<\/td>\n<td align=\"left\">\u7f13\u5b58\u4e00\u81f4\u6027\u534f\u8bae<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u672c\u8d28\u533a\u522b&#xff1a;<\/p>\n<ul>\n<li>CPU \u4e3a\u5ef6\u8fdf\u4f18\u5316&#xff1a;\u7528\u5927\u7f13\u5b58\u3001\u4e71\u5e8f\u6267\u884c\u3001\u5206\u652f\u9884\u6d4b&#xff0c;\u5c3d\u91cf\u8ba9\u5355\u6b21\u8bbf\u95ee\u5feb&#xff1b;<\/li>\n<li>GPU \u4e3a\u541e\u5410\u91cf\u4f18\u5316&#xff1a;\u5bb9\u5fcd\u9ad8\u5ef6\u8fdf&#xff0c;\u7528\u6d77\u91cf Warp \u5e76\u884c\u9690\u85cf\u5ef6\u8fdf&#xff0c;\u8ffd\u6c42\u5355\u4f4d\u65f6\u95f4\u5904\u7406\u66f4\u591a\u6570\u636e\u3002<\/li>\n<\/ul>\n<p>\u6570\u503c\u4f8b\u5b50&#xff1a;CPU \u8bbf\u95ee\u4e00\u6b21\u5185\u5b58 100 \u5468\u671f&#xff0c;\u4f46\u7f13\u5b58\u547d\u4e2d\u7387 95%&#xff0c;\u5e73\u5747\u5ef6\u8fdf <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">0.95\u00d75&#043;0.05\u00d7100\u22489.750.95 \\\\times 5 &#043; 0.05 \\\\times 100 \\\\approx 9.75<\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">0.95<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">5<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">0.05<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">100<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2248<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">9.75<\/span><\/span><\/span><\/span><\/span> \u5468\u671f\u3002<\/p>\n<p>GPU \u4e0d\u8ffd\u6c42\u5355\u6b21\u5feb&#xff0c;\u800c\u662f\u8ba9 64 \u4e2a Warp \u540c\u65f6\u8dd1&#xff0c;\u67d0\u4e2a Warp \u7b49\u5185\u5b58\u65f6\u5176\u4ed6 Warp \u7ee7\u7eed\u7b97\u3002<\/p>\n<h2>\u603b\u7ed3<\/h2>\n<table>\n<tr>\u5c42\u7ea7\u4f5c\u7528\u7f16\u7a0b\u63a7\u5236\u6570\u503c\u4f8b\u5b50\u4e2d\u7684\u89d2\u8272<\/tr>\n<tbody>\n<tr>\n<td align=\"left\">\u5168\u5c40\u5185\u5b58<\/td>\n<td align=\"left\">\u5b58\u6743\u91cd\u3001\u6570\u636e\u3001\u8f93\u51fa<\/td>\n<td align=\"left\">\u624b\u52a8 cudaMalloc<\/td>\n<td align=\"left\">\u4e0d\u5206\u5757\u65f6\u8bfb 21.4 \u4ebf\u6b21&#xff0c;\u5206\u5757\u540e\u8bfb 6700 \u4e07\u6b21<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">L2 \u7f13\u5b58<\/td>\n<td align=\"left\">\u7f13\u5b58\u70ed\u70b9\u3001\u8de8 SM \u5171\u4eab<\/td>\n<td align=\"left\">\u81ea\u52a8<\/td>\n<td align=\"left\">\u7f13\u5b58\u91cd\u590d\u8bbf\u95ee\u7684 A\u3001B \u5757<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5171\u4eab\u5185\u5b58<\/td>\n<td align=\"left\">Block \u5185\u534f\u4f5c\u3001\u5206\u5757<\/td>\n<td align=\"left\">\u624b\u52a8 __shared__<\/td>\n<td align=\"left\">\u77e9\u9635\u5206\u5757&#xff0c;\u5168\u5c40\u8bbf\u95ee\u51cf\u5c11 32 \u500d<\/td>\n<\/tr>\n<tr>\n<td align=\"left\">\u5bc4\u5b58\u5668<\/td>\n<td align=\"left\">\u7ebf\u7a0b\u5c40\u90e8\u53d8\u91cf<\/td>\n<td align=\"left\">\u7f16\u8bd1\u5668\u81ea\u52a8<\/td>\n<td align=\"left\">\u6bcf\u7ebf\u7a0b 24 \u4e2a\u5bc4\u5b58\u5668&#xff0c;\u52a0\u901f\u8ba1\u7b97<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>\u4e00\u53e5\u8bdd&#xff1a; GPU \u5185\u5b58\u6a21\u578b\u662f\u201c\u8fd1\u5feb\u5c0f\u3001\u8fdc\u6162\u5927\u201d\u7684\u56db\u5c42\u7ed3\u6784\u3002<\/p>\n<p>\u5bc4\u5b58\u5668 1 \u5468\u671f\u3001\u5171\u4eab\u5185\u5b58 33 \u5468\u671f\u3001L2 200 \u5468\u671f\u3001\u5168\u5c40\u5185\u5b58 290 \u5468\u671f\u3002\u77e9\u9635\u4e58\u6cd5\u4f8b\u5b50\u4e2d&#xff0c;\u4e0d\u5206\u5757\u76f4\u63a5\u8bfb\u5168\u5c40\u5185\u5b58\u7ea6 443 \u79d2&#xff0c;\u5206\u5757\u5230\u5171\u4eab\u5185\u5b58\u540e\u7ea6 39 \u79d2&#xff0c;\u52a0\u901f 11 \u500d\u3002GPU \u7528\u5206\u5c42\u5b58\u50a8 &#043; \u6d77\u91cf Warp \u5e76\u884c\u6765\u9690\u85cf\u9ad8\u5ef6\u8fdf&#xff0c;\u8ffd\u6c42\u541e\u5410\u91cf&#xff1b;CPU \u7528\u5927\u7f13\u5b58\u548c\u590d\u6742\u63a7\u5236\u6765\u964d\u4f4e\u5ef6\u8fdf\u3002<\/p>\n<p>\u7406\u89e3\u8fd9\u4e2a\u5c42\u6b21\u7ed3\u6784&#xff0c;\u662f\u5199\u51fa\u9ad8\u6027\u80fd GPU \u4ee3\u7801\u7684\u5173\u952e\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\/* MD \/ \u5bcc\u6587\u672c\u4e2d\u7684 .toc\uff08\u542b\u535a\u5ba2\u56ed\u642c\u5bb6\u7b49\u5d4c\u5957\u7ed3\u6784\uff09\uff1b.toc-box \u5728\u4fa7\u680f\uff0c\u4e0d\u53d7\u5f71\u54cd *\/#content_views .toc,\/* \u7f16\u8f91\u5668\u5e38\u5728\u76ee\u5f55\u524d\u540e\u63d2\u5165\u7a7a p\uff08:empty \u4ecd\u5360 20px\uff09\uff0c\u4e00\u5e76\u53bb\u6389\u907f\u514d\u9876\u7a7a\u9699 *\/#content_views.markdown_views &gt; p:empty:has(+ .toc),#content_views.markdown_views &gt; .toc + p:empty,\/* \u5bcc\u6587\u672c\u65e7\u7248\u76ee\u5f55\u6807\u8bb0 *\/#content<\/p>\n","protected":false},"author":2,"featured_media":122991,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[99,101,841,292,188],"topic":[],"class_list":["post-122992","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-server","tag-java","tag-spring","tag-transformer","tag-292","tag-188"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/122992.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\/* MD \/ \u5bcc\u6587\u672c\u4e2d\u7684 .toc\uff08\u542b\u535a\u5ba2\u56ed\u642c\u5bb6\u7b49\u5d4c\u5957\u7ed3\u6784\uff09\uff1b.toc-box \u5728\u4fa7\u680f\uff0c\u4e0d\u53d7\u5f71\u54cd *\/#content_views .toc,\/* \u7f16\u8f91\u5668\u5e38\u5728\u76ee\u5f55\u524d\u540e\u63d2\u5165\u7a7a p\uff08:empty \u4ecd\u5360 20px\uff09\uff0c\u4e00\u5e76\u53bb\u6389\u907f\u514d\u9876\u7a7a\u9699 *\/#content_views.markdown_views &gt; p:empty:has(+ .toc),#content_views.markdown_views &gt; .toc + p:empty,\/* \u5bcc\u6587\u672c\u65e7\u7248\u76ee\u5f55\u6807\u8bb0 *\/#content\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/122992.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-11-05T21:56:46+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/11\/20261105215645-6aecfb9d3cd63.png\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"4 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/122992.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/122992.html\",\"name\":\"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-11-05T21:56:46+00:00\",\"dateModified\":\"2026-11-05T21:56:46+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/122992.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/122992.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/122992.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/122992.html","og_locale":"zh_CN","og_type":"article","og_title":"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\/* MD \/ \u5bcc\u6587\u672c\u4e2d\u7684 .toc\uff08\u542b\u535a\u5ba2\u56ed\u642c\u5bb6\u7b49\u5d4c\u5957\u7ed3\u6784\uff09\uff1b.toc-box \u5728\u4fa7\u680f\uff0c\u4e0d\u53d7\u5f71\u54cd *\/#content_views .toc,\/* \u7f16\u8f91\u5668\u5e38\u5728\u76ee\u5f55\u524d\u540e\u63d2\u5165\u7a7a p\uff08:empty \u4ecd\u5360 20px\uff09\uff0c\u4e00\u5e76\u53bb\u6389\u907f\u514d\u9876\u7a7a\u9699 *\/#content_views.markdown_views &gt; p:empty:has(+ .toc),#content_views.markdown_views &gt; .toc + p:empty,\/* \u5bcc\u6587\u672c\u65e7\u7248\u76ee\u5f55\u6807\u8bb0 *\/#content","og_url":"https:\/\/www.wsisp.com\/helps\/122992.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-11-05T21:56:46+00:00","og_image":[{"url":"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2026\/11\/20261105215645-6aecfb9d3cd63.png"}],"author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"4 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/122992.html","url":"https:\/\/www.wsisp.com\/helps\/122992.html","name":"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-11-05T21:56:46+00:00","dateModified":"2026-11-05T21:56:46+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/122992.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/122992.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/122992.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"GPU \u5185\u5b58\u6a21\u578b\u6d45\u89e3 - \u56db\u5c42\u5b58\u50a8\u3001\u7f16\u7a0b\u63a7\u5236"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/122992","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=122992"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/122992\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media\/122991"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=122992"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=122992"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=122992"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=122992"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}