\u968f\u7740\u6211\u4eec\u63a8\u52a8\u7ecf\u5178\u8ba1\u7b97\u7684\u8fb9\u754c,\u9ad8\u6027\u80fd\u8ba1\u7b97\u7684\u672a\u6765\u770b\u8d77\u6765\u66f4\u52a0\u4ee4\u4eba\u5174\u594b\uff1a<\/p>\n

\n
• \u767e\u4ebf\u4ebf\u6b21\u7ea7\u8ba1\u7b97\uff1a<\/strong>\u80fd\u591f\u6bcf\u79d2\u8fdb\u884c\u4e00\u767e\u4ebf\u4ebf\u6b21(10^18)\u8ba1\u7b97\u7684\u7cfb\u7edf<\/li>\n
• \u91cf\u5b50HPC\uff1a<\/strong>\u5229\u7528\u91cf\u5b50\u529b\u5b66\u5b9e\u73b0\u4ee4\u4eba\u96be\u4ee5\u7f6e\u4fe1\u7684\u8ba1\u7b97\u80fd\u529b<\/li>\n
• AI-HPC\u878d\u5408\uff1a<\/strong>\u6df1\u5ea6\u5b66\u4e60\u7b97\u6cd5\u4f18\u5316\u5de5\u4f5c\u6d41\u7a0b,\u53cd\u4e4b\u4ea6\u7136<\/li>\n<\/ul>\n

  [\/vc_column_text][\/vc_column][\/vc_row][vc_row el_class=”blog-detail-section”][vc_column][vc_column_text]<\/p>\n

  \u62e5\u62b1HPC\u9769\u547d<\/strong><\/h2>\n

  \u9ad8\u6027\u80fd\u8ba1\u7b97\u4e0d\u4ec5\u4ec5\u662f\u5173\u4e8e\u539f\u59cb\u529f\u7387 – \u5b83\u662f\u5173\u4e8e\u63a8\u52a8\u8ba1\u7b97\u53ef\u80fd\u6027\u7684\u8fb9\u754c\u3002\u5f53\u6211\u4eec\u8fdb\u5165\u767e\u4ebf\u4ebf\u6b21\u7ea7\u8ba1\u7b97\u65f6\u4ee3\u53ca\u66f4\u8fdc\u7684\u672a\u6765,HPC\u5c06\u7ee7\u7eed\u5904\u4e8e\u79d1\u5b66\u53d1\u73b0\u548c\u6280\u672f\u521b\u65b0\u7684\u524d\u6cbf\u3002\u65e0\u8bba\u60a8\u662f\u5728\u6a21\u62df\u5b87\u5b99\u8fd8\u662f\u8bad\u7ec3\u4e0b\u4e00\u4ee3AI\u6a21\u578b,HPC\u90fd\u63d0\u4f9b\u4e86\u5c06\u60a8\u6700\u75af\u72c2\u7684\u6280\u672f\u68a6\u60f3\u53d8\u4e3a\u73b0\u5b9e\u7684\u8ba1\u7b97\u80fd\u529b\u3002<\/p>\n

  [\/vc_column_text][\/vc_column][\/vc_row]<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":"

  [vc_row el_class=”blog-detail-section”][vc_column][vc_column_text] \u5728\u8ba1\u7b97\u80fd\u529b\u7684\u9886\u57df\u4e2d,\u9ad8\u6027\u80fd\u8ba1\u7b97(HPC)\u72ec\u5360\u9ccc\u5934\u3002\u672c\u6587\u6df1\u5165\u63a2\u8ba8HPC\u7684\u590d\u6742\u6027,\u63a2\u7d22\u5176\u6838\u5fc3\u7ec4\u4ef6\u3001\u5e94\u7528\u548c\u6240\u91ca\u653e\u7684\u539f\u59cb\u8ba1\u7b97\u80fd\u529b\u3002\u65e0\u8bba\u60a8\u662f\u72ec\u7acb\u670d\u52a1\u5668\u7528\u6237\u8fd8\u662f\u597d\u5947\u7684\u6280\u672f\u7231\u597d\u8005,\u51c6\u5907\u6df1\u5165\u63a2\u7d22\u5e76\u884c\u5904\u7406\u3001GPU\u52a0\u901f\u548c\u4ee4\u4eba\u60ca\u53f9\u7684\u8ba1\u7b97\u58ee\u4e3e\u7684\u4e16\u754c\u5427\u3002 [\/vc_column_text][\/vc_column][\/vc_row][vc_row el_class=”blog-detail-section”][vc_column][vc_column_text] \u89e3\u7801HPC\uff1a\u8d85\u8d8a\u666e\u901a\u7684\u6570\u5b57\u8ba1\u7b97\u5668 HPC\u7684\u6838\u5fc3\u662f\u5229\u7528\u6d77\u91cf\u8ba1\u7b97\u8d44\u6e90\u6765\u89e3\u51b3\u590d\u6742\u95ee\u9898\u3002\u4e0e\u6807\u51c6\u53f0\u5f0f\u673a\u751a\u81f3\u5f3a\u5927\u7684\u5de5\u4f5c\u7ad9\u4e0d\u540c,\u5176\u7cfb\u7edf\u8bbe\u8ba1\u7528\u4e8e\u5904\u7406\u666e\u901a\u8ba1\u7b97\u673a\u96be\u4ee5\u5e94\u5bf9\u7684\u4efb\u52a1\u3002\u60f3\u8c61\u6a21\u62df\u661f\u7cfb\u7684\u8bde\u751f\u3001\u9884\u6d4b\u5168\u7403\u5929\u6c14\u6a21\u5f0f\u6216\u63ed\u5f00\u86cb\u767d\u8d28\u6298\u53e0\u7684\u5965\u79d8 – \u8fd9\u5c31\u662fHPC\u5927\u663e\u8eab\u624b\u7684\u5730\u65b9\u3002 [\/vc_column_text][\/vc_column][\/vc_row][vc_row el_class=”blog-detail-section”][vc_column][vc_column_text] \u8ba1\u7b97\u5de8\u517d\u7684\u6784\u5efa\u6a21\u5757 \u9ad8\u6027\u80fd\u8ba1\u7b97\u7cfb\u7edf\u662f\u5c16\u7aef\u786c\u4ef6\u7ec4\u4ef6\u5b8c\u7f8e\u548c\u8c10\u5de5\u4f5c\u7684\u4ea4\u54cd\u4e50\uff1a CPU\uff1a\u591a\u6838\u91ce\u517d,\u5177\u6709\u9ad8\u65f6\u949f\u901f\u5ea6\u548c\u5de8\u5927\u7f13\u5b58 GPU\uff1a\u5e76\u884c\u5904\u7406\u5f3a\u8005,\u52a0\u901f\u7279\u5b9a\u8ba1\u7b97 \u5185\u5b58\uff1a\u5927\u91cf\u9ad8\u5e26\u5bbd\u3001\u4f4e\u5ef6\u8fdf\u7684RAM \u5b58\u50a8\uff1a\u95ea\u7535\u822c\u5feb\u901f\u7684SSD\u548c\u7528\u4e8e\u6570\u636e\u5bc6\u96c6\u578b\u64cd\u4f5c\u7684\u5e76\u884c\u6587\u4ef6\u7cfb\u7edf \u4e92\u8fde\uff1a\u5982InfiniBand\u7b49\u9ad8\u5e26\u5bbd\u7f51\u7edc,\u7528\u4e8e\u8282\u70b9\u95f4\u65e0\u7f1d\u901a\u4fe1 [\/vc_column_text][\/vc_column][\/vc_row][vc_row el_class=”blog-detail-section”][vc_column][vc_column_text] \u5e76\u884c\u5904\u7406\uff1a\u5206\u800c\u6cbb\u4e4b HPC\u7684\u79d8\u8bc0\u5728\u4e8e\u5e76\u884c\u5904\u7406\u3002\u901a\u8fc7\u5c06\u590d\u6742\u95ee\u9898\u5206\u89e3\u6210\u66f4\u5c0f\u7684\u5757,\u5e76\u5728\u6570\u767e\u6216\u6570\u5343\u4e2a\u8ba1\u7b97\u8282\u70b9\u4e0a\u540c\u65f6\u5904\u7406,HPC\u7cfb\u7edf\u5b9e\u73b0\u4e86\u4ee4\u4eba\u96be\u4ee5\u7f6e\u4fe1\u7684\u901f\u5ea6\u3002\u4ee5\u4e0b\u662f\u4f7f\u7528Python\u7684multiprocessing\u6a21\u5757\u8fdb\u884c\u5e76\u884c\u5904\u7406\u7684\u7b80\u5355\u793a\u4f8b\uff1a import multiprocessing as mp import math def calculate_pi(n): h = 1.0 \/ n s = 0.0 for i in range(n): x = h * (i + 0.5) s += 4.0 \/ (1.0 + […]<\/p>\n

  Read More…<\/a><\/p>\n","protected":false},"author":2,"featured_media":15200,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[62],"tags":[3307,3308,3309,3310,1734],"class_list":["post-15239","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-hong-kong-dedicated-server-sc","tag-hpc-sc","tag-parallel-processing-sc","tag-scientific-computing-sc","tag-gpu-acceleration-sc","tag-high-performance-computing-sc"],"acf":[],"yoast_head":"\n