基於機器學習的分子動力學模擬獲得戈登·貝爾獎

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"2020年"},{"type":"link","attrs":{"href":"https:\/\/www.acm.org\/","title":"","type":null},"content":[{"type":"text","text":"美國計算機協會"}]},{"type":"text","text":"（ACM）的"},{"type":"link","attrs":{"href":"https:\/\/www.acm.org\/media-center\/2020\/november\/gordon-bell-prize-2020","title":"","type":null},"content":[{"type":"text","text":"戈登·貝爾獎"}]},{"type":"text","text":"授予了來自美國和中國機構的研究團隊，表彰他們題爲“用機器學習將分子動力學的從頭計算方法的精度推至1億個原子”的項目。據團隊介紹，深勢分子動力學（Deep Potential Molecular dynamics，DPMD）是一種基於機器學習的新協議，它能夠每天模擬超過1億個原子的1納米的軌跡。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"http:\/\/www.sciencedirect.com\/topics\/biochemistry-genetics-and-molecular-biology\/molecular-dynamics","title":"","type":null},"content":[{"type":"text","text":"分子動力學（Molecular Dynamics）"}]},{"type":"text","text":"是一種計算機模擬方法，用來分析在特定的時間段內原子的運動和相互作用。從小到單細胞的系統，到大到氣體雲的複雜系統，科學家都能利用分子動力學模擬的方式來了解這些分子化合物在一段時間內的行動。三十五年來，研究人員一直在使用一種被稱爲"},{"type":"link","attrs":{"href":"https:\/\/www.pnas.org\/content\/102\/19\/6654","title":"","type":null},"content":[{"type":"text","text":"從頭計算（ab initio）的模擬方法進行分子動力學研究"}]},{"type":"text","text":"，因爲事實證明這種方式是最有效的。儘管從頭計算方法（在拉丁語中，它的意思是“從第一原理”）在模擬中取得了很高的精確度，但是這種方式需要大量的計算資源，這限制了它的應用，最多隻能用於包含數千個原子的小型系統。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"DPMD背後的團隊在"},{"type":"link","attrs":{"href":"https:\/\/dl.acm.org\/doi\/pdf\/10.5555\/3433701.3433707","title":"","type":null},"content":[{"type":"text","text":"本論文"}]},{"type":"text","text":"中詳細介紹了“從頭計算”方法的侷限性，發現它隨電子自由度（electronic degrees of freedom）的數量呈立方級關係。採用從頭計算方法可以實現的典型時空比例的設置是~100個原子和~10皮秒。從頭計算方法幾乎完美地遵守立方擴展定律。即便是世界上最大的超級計算機，也無法進行復雜的化學反應、電化學電池、納米晶體材料和輻射損傷等方面的模擬。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"blockquote","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"DP（深度勢能，Deep Potential）模型的精確性來源於"},{"type":"link","attrs":{"href":"https:\/\/www.nextplatform.com\/2020\/04\/07\/changing-conditions-for-neural-network-processing\/","title":"","type":null},"content":[{"type":"text","text":"深度神經網絡（DNN）"}]},{"type":"text","text":"對高維函數獨特的近似能力、對對稱性約束等物理要求的恰當處理以及併發的學習模式，能夠生成緊湊的訓練數據集，並保證在相關配置空間中統一的精確度。"}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"DPMD團隊選擇利用世界第二快的超級計算機"},{"type":"link","attrs":{"href":"https:\/\/www.ibm.com\/thought-leadership\/summit-supercomputer\/","title":"","type":null},"content":[{"type":"text","text":"IBM的Summit系統"}]},{"type":"text","text":"上的GPU來運行幾乎所有的計算和通信任務。由於“深度勢能”模型中的計算粒度的限制，該團隊發現，僅僅嚴重依賴GPU的效率會很低。通過算法創新，包括爲相鄰列表提供新的數據佈局以避免嵌入式矩陣計算中的分支，將新數據結構中的元素壓縮爲64位整數以提高GPU對自定義"},{"type":"link","attrs":{"href":"https:\/\/www.tensorflow.org\/","title":"","type":null},"content":[{"type":"text","text":"TensorFlow"}]},{"type":"text","text":"操作的優化，以及爲深度勢能模型創建混合精度計算，團隊針對GPU相關的低效率進行了優化。通過這些改進，研究人員能夠以從頭計算計算相同的精度模擬前所未有的規模和時間範圍。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"戈登·貝爾獎旨在表彰高性能計算領域的成就，入圍者必須要證明他們的算法能夠在世界最強大的超級計算機上進行擴展。GPU Deep MD-Kit能夠有效地擴展到整個Summit超級計算機上，在單\/半混合精度下達到91 PFLOPS（一個PFLOPS指的是每秒1千萬億次的浮點運算——譯註）和162\/275 PFLOPS。這一成績爲下一代超級計算機更好地實現機器學習和物理建模的結合提出了新的挑戰。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"原文鏈接："}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"link","attrs":{"href":"https:\/\/www.infoq.com\/news\/2020\/12\/ml-based-molecular-dynamics\/","title":"","type":null},"content":[{"type":"text","text":"Molecular Dynamics Simulation Based on Machine Learning Wins Gordon Bell Prize"}]}]}]}