Domain‐specific virtual processors as a portable programming and execution model for parallel computational workloads on modern heterogeneous high‐performance computing architectures

Lyakh Dmitry I.

摘要

Abstract >We advocate domain‐specific virtual processors (DSVP) as a portability layer for expressing and executing domain‐specific computational workloads on modern heterogeneous HPC architectures, with applications in quantum chemistry. Specifically, in this article we extend, generalize and better formalize the concept of a domain‐specific virtual processor as applied to scientific high‐performance computing. In particular, we introduce a system‐wide recursive (hierarchical) hardware encapsulation mechanism into the DSVP architecture and specify a concrete microarchitectural design of an abstract DSVP from which specialized DSVP implementations can be derived for specific scientific domains. Subsequently, we demonstrate, an example of a domain‐specific virtual processor specialized to numerical tensor algebra workloads, which is implemented in the ExaTENSOR library developed by the author with a primary focus on the quantum many‐body computational workloads on large‐scale GPU‐accelerated HPC platforms. </abstract> </span> <span class="z_kbtn z_kbtnclass hoverxs" style="display: none;">展开▼</span> </div> <div class="translation abstracttxt"> <span class="zhankaihshouqi fivelineshidden" id="abstract"> <span>机译：</span><Abstract Type =“Main”XML：Lang =“en”> <标题类型=“main”>抽象</ title> >我们将特定于域的虚拟处理器（DSVP）提升为用于表达和执行域的可移植性图层在现代异质HPC架构上的特定计算工作负载，具有量子化学应用。具体地，在本文中，我们扩展，概括和更好地将域特定虚拟处理器的概念正式化，应用于科学的高性能计算。特别是，我们将系统范围的递归（分层）硬件封装机制介绍到DSVP架构中，并指定了一个抽象DSVP的混凝土微架构设计，可以从中导出专门的DSVP实现，用于特定的科学域。随后，我们演示了专用于数值张量代数工作负载的域特定虚拟处理器的示例，该工作负载在由作者开发的延长句子库中实现，主要专注于大规模GPU上的量子数量计算工作负载 - 加速HPC平台。</ p> </ abstract> </span> <span class="z_kbtn z_kbtnclass hoverxs" style="display: none;">展开▼</span> </div> </div> <div class="record"> <h2 class="all_title" id="enpatent33" >著录项</h2> <ul> <li> <span class="lefttit">来源</span> <div style="width: 86%;vertical-align: text-top;display: inline-block;"> <a href='/journal-foreign-21753/'>《International Journal of Quantum Chemistry》</a> <b style="margin: 0 2px;">|</b><span>2019年第12期</span><b style="margin: 0 2px;">|</b><span>共19页</span> </div> </li> <li> <div class="author"> <span class="lefttit">作者</span> <p id="fAuthorthree" class="threelineshidden zhankaihshouqi"> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Lyakh Dmitry I.&option=202" target="_blank" rel="nofollow">Lyakh Dmitry I.;</a> </p> <span class="z_kbtnclass z_kbtnclassall hoverxs" id="zkzz" style="display: none;">展开▼</span> </div> </li> <li> <div style="display: flex;"> <span class="lefttit">作者单位</span> <div style="position: relative;margin-left: 3px;max-width: 639px;"> <div class="threelineshidden zhankaihshouqi" id="fOrgthree"> <p>Scientific Computing National Center for Computational SciencesOak Ridge National LaboratoryOak Ridge Tennessee;</p> </div> <span class="z_kbtnclass z_kbtnclassall hoverxs" id="zhdw" style="display: none;">展开▼</span> </div> </div> </li> <li > <span class="lefttit">收录信息</span> <span style="width: 86%;vertical-align: text-top;display: inline-block;"></span> </li> <li> <span class="lefttit">原文格式</span> <span>PDF</span> </li> <li> <span class="lefttit">正文语种</span> <span>eng</span> </li> <li> <span class="lefttit">中图分类</span> <span><a href="https://www.zhangqiaokeyan.com/clc/1186.html" title="物理化学（理论化学）、化学物理学">物理化学（理论化学）、化学物理学;</a></span> </li> <li class="antistop"> <span class="lefttit">关键词</span> <p style="width: 86%;vertical-align: text-top;"> <a style="color: #3E7FEB;" href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=heterogeneous node architecture&option=203" rel="nofollow">heterogeneous node architecture;</a> <a style="color: #3E7FEB;" href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=high‐performance computing&option=203" rel="nofollow">high‐performance computing;</a> <a style="color: #3E7FEB;" href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=numerical tensor algebra&option=203" rel="nofollow">numerical tensor algebra;</a> <a style="color: #3E7FEB;" href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=scientific computing&option=203" rel="nofollow">scientific computing;</a> <a style="color: #3E7FEB;" href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=virtual machine&option=203" rel="nofollow">virtual machine;</a> </p> <div class="translation"> 机译：异构节点架构;高性能计算;数值张量代数;科学计算;虚拟机; </div> </li> </ul> </div> </div> <div class="literature cardcommon"> <div class="similarity "> <h3 class="all_title" id="enpatent66">相似文献</h3> <div class="similaritytab clearfix"> <ul> <li class="active" >外文文献</li> <li >中文文献</li> <li >专利</li> </ul> </div> <div class="similarity_details"> <ul > <li> <div> <b>1. </b><a class="enjiyixqcontent" href="/journal-foreign-detail/0704019755376.html">Domain‐specific virtual processors as a portable programming and execution model for parallel computational workloads on modern heterogeneous high‐performance computing architectures</a> <b>[J]</b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Lyakh Dmitry I.&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Lyakh Dmitry I. </a> <a href="/journal-foreign-21753/" target="_blank" rel="nofollow" class="tuijian_authcolor">International Journal of Quantum Chemistry .</a> <span>2019</span><span>,第12期</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：域特定的虚拟处理器作为现代异构高性能计算架构上并行计算工作负载的便携式编程和执行模型</span> </p> </li> <li> <div> <b>2. </b><a class="enjiyixqcontent" href="/academic-journal-foreign_experimental-mechanics_thesis/0204119851256.html">Performance portable parallel programming of heterogeneous stencils across shared-memory platforms with modern Intel processors</a> <b>[J]</b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Szustak Lukasz&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Szustak Lukasz,</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Bratek Pawel&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Bratek Pawel </a> <a href="/journal-foreign-2024/" target="_blank" rel="nofollow" class="tuijian_authcolor">Experimental Mechanics .</a> <span>2019</span><span>,第3期</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：具有现代英特尔处理器的共享内存平台异构模板的性能平行编程</span> </p> </li> <li> <div> <b>3. </b><a class="enjiyixqcontent" href="/academic-journal-foreign_journal-parallel-distributed-computing_thesis/0204113013484.html">Performance portability study for massively parallel computational fluid dynamics application on scalable heterogeneous architectures</a> <b>[J]</b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Lee Seyong&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Lee Seyong,</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Gounley John&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Gounley John,</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Randles Amanda&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Randles Amanda,</a> <a href="/journal-foreign-2082/" target="_blank" rel="nofollow" class="tuijian_authcolor">Journal of Parallel and Distributed Computing .</a> <span>2019</span><span>,第JULa期</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：可扩展异构架构上大规模并行计算流体动力学应用的性能可移植性研究</span> </p> </li> <li> <div> <b>4. </b><a class="enjiyixqcontent" href="/academic-conference-foreign_meeting-224819_thesis/020516282127.html">NUMERICAL REPRODUCIBILITY, PORTABILITY AND PERFORMANCE OF MODERN PSEUDO RANDOM NUMBER GENERATORS: PRELIMINARY STUDY FOR PARALLEL STOCHASTIC SIMULATIONS USING HYBRID XEON PHI COMPUTING PROCESSORS</a> <b>[C]</b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Van Toan Dao&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Van Toan Dao,</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Hong Quang Nguyen&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Hong Quang Nguyen,</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Lydia Maigne&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Lydia Maigne,</a> <a href="/conference-foreign-224819/" target="_blank" rel="nofollow" class="tuijian_authcolor">European simulation and modelling conference .</a> <span>2014</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：现代伪随机数发生器的数值可再现性，可移植性和性能：使用混合氙气PHI计算处理器进行并行随机模拟的初步研究</span> </p> </li> <li> <div> <b>5. </b><a class="enjiyixqcontent" href="/academic-degree-foreign_mphd_thesis/02061850595.html">Performance Analysis of Program Executions on Modern Parallel Architectures</a> <b>[D] </b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Liu, Xu.&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Liu, Xu. </a> <span>2014</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：现代平行架构方案执行性能分析</span> </p> </li> <li> <div> <b>6. </b><a class="enjiyixqcontent" href="/academic-journal-foreign-pmc_detail_thesis/040007287992.html">Parallelisation of equation-based simulation programs on heterogeneous computing systems</a> <b>[O] </b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Dragan D. Nikolić&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Dragan D. Nikolić </a> <span>2018</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：基于等式的异构计算系统仿真程序的平行化</span> </p> </li> <li> <div> <b>7. </b><a class="enjiyixqcontent" href="/open-access_resources_thesis/0100063380959.html ">Two-Dimensional Matrix Partitioning for Parallel Computing on Heterogeneous Processors Based on Their Functional Performance Models</a> <b>[O] </b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Alexey Lastovetsky&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Alexey Lastovetsky,</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Ravi Reddy&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Ravi Reddy </a> <span>2015</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：基于功能性能模型的异构处理器并行计算的二维矩阵划分</span> </p> </li> <li> <div> <b>8. </b><a class="enjiyixqcontent" href="/ntis-science-report_ad_thesis/020711492759.html">Distributed Computing for Signal Processing: Modeling of Asynchronous Parallel Computation. Appendix G. On the Design and Modeling of Special Purpose Parallel Processing Systems</a> <b>[R] </b> . <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=Smith, B. W.&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">Smith, B. W. </a> <span>1985</span> </span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：信号处理的分布式计算：异步并行计算的建模。附录G.关于专用并行处理系统的设计和建模</span> </p> </li> </ul> <ul style="display: none;"> <li> <div> <b>1. </b><a class="enjiyixqcontent" href="/academic-journal-cn_computer-engineering-science_thesis/0201240952171.html">主从式单边异构多核处理器编程模型和编译架构</a> <b>[J]</b> <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=李春江&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor"> . 李春江</a> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=杨学军&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor">,杨学军</a> <span> <a href="/journal-cn-8895/" target="_blank" rel="nofollow" class="tuijian_authcolor"> . 计算机工程与科学 </a> </span> <span> . 2009</span><span>,第008期</span> </span> </div> </li> <li> <div> <b>2. </b><a class="enjiyixqcontent" href="/academic-degree-domestic_mphd_thesis/020315127432.html">混合异构架构上的生物大数据并行计算研究</a> <b>[A] </b> <span> <a href="/search.html?doctypes=4_5_6_1-0_4-0_1_2_3_7_9&sertext=兰海东&option=202" target="_blank" rel="nofollow" class="tuijian_auth tuijian_authcolor"> . 兰海东</a> <span> . 2018</span> </span> </div> </li> </ul> <ul style="display: none;"> <li> <div> <b>1. </b><a class="enjiyixqcontent" href="/patent-detail/061203094181.html">用于实现代码划分和在异构处理器内核上执行的高效有向非循环图模式匹配</a> <b>[P]</b> . <span> 中国专利： CN105474172B </span> <span> . 2018.11.09</span> </div> </li> <li> <div> <b>2. </b><a class="enjiyixqcontent" href="/patent-detail/06120107953884.html">用于实现代码划分和在异构处理器内核上执行的高效有向非循环图模式匹配</a> <b>[P]</b> . <span> 中国专利： CN105474172A </span> <span> . 2016-04-06</span> </div> </li> <li> <div> <b>3. </b><a class="enjiyixqcontent" href="/patent-detail/06130406833066.html">Computation method for computing performance value when processor executes code, computation apparatus for computing performance value when processor executes code, and computer readable storage medium storing computation program for computing performance value when processor executes code</a> <b>[P]</b> . <span> 外国专利：  US9880841B2 </span> <span> . 2018-01-30</span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：用于在处理器执行代码时计算性能值的计算方法，用于在处理器执行代码时计算性能值的计算装置以及存储用于在处理器执行代码时计算性能值的计算程序的计算机可读存储介质 </span> </p> </li> <li> <div> <b>4. </b><a class="enjiyixqcontent" href="/patent-detail/06130404526029.html">Creating models based on performance metrics of a computing workloads running in a plurality of data centers to distribute computing workloads</a> <b>[P]</b> . <span> 外国专利：  US10346204B2 </span> <span> . 2019-07-09</span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：基于在多个数据中心中运行的计算工作负载的性能指标创建模型以分配计算工作负载 </span> </p> </li> <li> <div> <b>5. </b><a class="enjiyixqcontent" href="/patent-detail/06130420844138.html">METHODS AND SYSTEMS FOR PRICING COST OF EXECUTION OF A PROGRAM IN A PARALLEL PROCESSING ENVIRONMENT AND FOR AUCTIONING COMPUTING RESOURCES FOR EXECUTION OF PROGRAMS</a> <b>[P]</b> . <span> 外国专利：  US2013226724A1 </span> <span> . 2013-08-29</span> </div> <p class="zwjiyix translation" style="max-width: initial;height: auto;word-break: break-all;white-space: initial;text-overflow: initial;overflow: initial;"> <span>机译：在并行处理环境中对程序执行成本进行定价并为程序执行对计算资源进行拍卖的方法和系统 </span> </p> </li> </ul> </div> </div> </div> <div class="theme cardcommon" style="overflow: auto;display:none"> <h3 class="all_title" id="enpatent55">相关主题</h3> <ul id="subject"> </ul> </div> </div> </div> </div> <div class="right rightcon"> <div class="details_img cardcommon clearfix" style="margin-bottom: 10px;display:none;" > </div> </div> </div> <div id="thesis_get_original1" class="downloadBth" style="bottom: 19px;z-index: 999;" onclick="ywcd('0704019755376','4',7,2,1,'',this,24)" class="delivery" prompt="010401" title="通过人工服务将文献原文发送至邮箱" >获取原文</div> <div class="journalsub-pop-up" style="display: none"> <div class="journal-sub"> <h2>期刊订阅</h2> <img src="https://cdn.zhangqiaokeyan.com/img/loginclose.png" alt="关闭" onclick="$('.journalsub-pop-up').hide()"> <p class="pardon">抱歉，该期刊暂不可订阅，敬请期待！</p> <p class="current">目前支持订阅全部北京大学中文核心（2020）期刊目录。</p> <div style="display: flex;margin-top: 69px;justify-content: space-between;"> <div class="no-sub" onclick="$('.journalsub-pop-up').hide()">暂不订阅</div> <div class="other-sub" onclick="continueSub('from=pc-detail')">继续订阅其他期刊</div> </div> </div> </div> <div class="right_btn"> <ul> <li class="gouwuche">  </li> <li class="yijian"> <a href="javascript:void(0);" onclick="link_analysis('/mycenter/auth/complaint.html',this)" title="意见反馈">意见反馈</a> </li> <li class="top"> <a href="javascript:scrollTo(0,0);" title="回到顶部">回到顶部</a> </li> <li class="shouye"> <a href="/" title="首页">回到首页</a> </li> </ul> </div> <div class="xllindexfooter"> <div class="xllindexfootercenter"> <div class="xllindexfooterleft left" > <div class="xllindexfooterleftli"> <ul> <li><a href="/about.html" title="关于掌桥">关于掌桥</a></li> <li><a href="/help/helpmap.html" title="资源导航">资源导航</a></li> <li><a href="/help/helpguide.html" title="新手指南">新手指南</a></li> <li><a href="/help/helpcenter.html" title="常见问题">常见问题</a></li> <li><a href="/sitemap.html" title="网站地图">网站地图</a></li> <li><a href="/help/helpcenter.html?type=9" title="版权声明">版权声明</a></li> </ul> </div> <div class="xllindexfooterleft"> <p class="xllindexfooterlefteamil">客服邮箱：kefu@zhangqiaokeyan.com</p> <div class="xllindexfooterlefttcp"> <div class="xllindexfooterpoliceiimg"></div> <div class="xllindexfooterpoliceispan"> <span>京公网安备：11010802029741号 </span> <span>ICP备案号：<a href="https://beian.miit.gov.cn" rel="nofollow" target="_blank" title="ICP备案号">京ICP备15016152号-6</a></span> <span>六维联合信息科技 (北京) 有限公司©版权所有</span> </div> </div> </div> </div> <div class="xllindexfooterright left"> <ul> <li> <p style="font-weight: bold;">客服微信</p> <div></div> </li> <li> <p style="font-weight: bold;">服务号</p> <div></div> </li> </ul> </div> </div> </div> <span id="0704019755376down" data-source="7," data-out-id="x1sv3gzBkpCTUXNobAX8buswxioy2VYNV8iacvrCrA12nC3VzSa7S39WFKQ1g1bv," data-f-source-id="7" data-title="Domain‐specific virtual processors as a portable programming and execution model for parallel computational workloads on modern heterogeneous high‐performance computing architectures" data-price="20" data-site-name="" data-transnum="24" style="display:none;"></span> <input type="hidden" value="4" id="sourcetype"> <input type="hidden" value="21753" id="journalid"> <input type="hidden" value="1" id="inyn_provide_service_level"> <input type="hidden" value="https://cdn.zhangqiaokeyan.com" id="imgcdn"> <input type="hidden" value="1" id="isdeatail"> <input type="hidden" value="" id="syyn_indexed_database"> <input type="hidden" value="" id="servicetype"> <input type="hidden" id="pagename" value="Domain‐specific virtual processors as a portable programming and execution model for parallel computational workloads on modern heterogeneous high‐performance computing architectures"/> <input type="hidden" value="thesis_get_original" id="pageIdentification"> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/jquery-1.12.4.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/lwlh_ajax.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/zq.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/common.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/jquery.cookie.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/top.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/tip.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/login.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/down.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/search.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/newmail.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/searchtype.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/user/regist.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/zxs_solor/detail.js?v=5.7.8"></script> <script type="text/javascript" src="https://www.zhangqiaokeyan.com/statistics/static/pagecollection.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/tj.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/sse.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/pushbaidu.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/common/history.js"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/util/cookie.js?v=5.7.8"></script> <script type="text/javascript" src="https://cdn.zhangqiaokeyan.com/js/weipu/weipu.js?v=5.7.8"></script> </body> <script> $(function(){ var weiPuStatus = getCookie('WeiPuStatus'); if(weiPuStatus){ tipWeiPuStatus(weiPuStatus); delCookie('WeiPuStatus'); } getFacetKeywordVoInId(); var sourcetype = $("#sourcetype").val(); var inyn_provide_service_level = $("#inyn_provide_service_level").val(); if((sourcetype ==1||sourcetype==4)&&inyn_provide_service_level!=4){ getJournal(); } }) </script> </html>