View: China set aside for a "super computer power" itself-super computer, supercomputer-IT information
New issue of the world's TOP500 Supercomputers list published on June 20, overnight became the world's first Super computer in China the biggest winner on the stage: using China's own chip-making "shenwei ? Bank of light" instead of "Tianhe second" top; and China's supercomputer list surpassed the total United States ranked first. In addition, China Green500 supercomputer performance rankings in the top 10 of the list, also accounted for half, including "power light ? Taihu", medium wave 1, Association dawn 3, 1.
One of the biggest surprises is "power light ? Taihu". Chinese researchers produced almost two generations behind the technology with Intel's next generation of nuclear products, also a step ahead than the latter application. Does not use Intel chips, not idle waste of computing resources criticized "power light ? Taihu" ascent is almost Chinese supercomputer on United States supercomputer a perfect counter attack.
Today, China can use: supercomputer power itself? In my opinion, not to heat. Sponsor of the TOP500 global supercomputer jieke·tangjiala said, supercomputers, to as simple and easy as possible. This is his best (Super) definition of your computer. Dongarra noted that the supercomputer industry cannot avoid the question, that is, practical.
Former Chinese "Tianhe second" climb to the top of the TOP500 era, Chinese supercomputer has been questioned like this: Chinese supercomputer is good? If excess? Ever good?
China does have weaknesses on the super computer. First of all is the lack of application software, many applications rely on commercial software, relatively few active research and development, followed by lack of super computer expertise; is the lack of effective cooperation in the super computer in China again.
Super computer equipment hardware and system are all professional people to do professional things, but their application environment is interdisciplinary. If the lack of effective cooperation, application development "hard stuff" no one wants to eat. In addition, the application environment more conducive to high performance computing users, and also large bottleneck in the development of high-performance computing.
Super computer personnel training is also an important part of building a supercomputer eco. The next 10 years, super computer technology will develop by leaps and bounds and petaflops supercomputers in the future volume may only candy bar the size of their power, stability and other technical challenges still depend on the younger generation to solve. In ordinary higher education in China, there is no "parallel computing" this profession.
After all, we need to improve the ecological environment of high performance computing, supercomputer industry healthy, sustainable development can be achieved. "Power light ? Taihu" surprise for China, and China on a selected number of "spot" it gave us confidence and courage in the supercomputer field, but has not been produced before the health of the environment, we might as well not to "super computer power" itself. Rather than practical thing to do right now, accumulation of power, waiting for ripe on that day.
观点：中国暂且不要以“超级计算机第一强国”自居 - 超级计算机,超级计算机 - IT资讯
新一期全球超级计算机TOP500榜单6月20日公布，中国一夜之间成为世界超级计算机舞台上的最大赢家：使用中国自主芯片制造的“神威·太湖之光”取代“天河二号”登上榜首；且中国超级计算机上榜总数首次超越美国名列第一。此外，中国超级计算机在效能排行榜Green500榜单的前10名中，也占据了半壁江山，包括“神威·太湖之光”、中科曙光3台、浪潮1台、联想1台。
其中最大的惊喜当属“神威·太湖之光”。中国科研人员用几乎落后两代的工艺制造出能与英特尔下一代众核相当的产品，还比后者先一步投入应用。没有使用英特尔芯片、没有计算资源闲置空耗的诟病，“神威·太湖之光”的登顶几乎是中国超级计算机对美国超级计算机的一次完美逆袭。
那么，今天中国可以以超级计算机第一强国自居了吗？笔者认为，还没到火候。全球超级计算机TOP500的发起人杰克·唐加拉曾说，超级计算机要尽量简单好用。这是他对最好（超级）计算机的定义。唐加拉指出了超级计算机行业无法回避的问题，即实用性。
在此前中国“天河二号”登顶TOP500的时代，中国超级计算机一直遭受这样的质疑：中国超级计算机是否好用？是否过剩？有没有用好？
中国在超级计算机应用上的确有弱点。首先是缺乏应用软件，很多应用依赖商业软件，主动研发相对较少；其次是缺乏大量的超级计算机专门人才；再次是中国在超级计算机应用中缺乏有效合作。
超级计算机设备的硬件和系统都是专业的人做专业的事，但其应用环境却是跨学科的。如果缺乏有效合作，应用开发的“硬骨头”就没人愿意去啃。此外，把高性能计算机的应用环境做得更利于用户使用，也是目前高性能计算发展中较大的瓶颈。
超级计算机人才的培养也是构建超级计算机生态的重要一环。未来10年，超级计算机技术还会有更加突飞猛进的发展，未来千万亿次超级计算机的体积可能只有糖块大小，其面临的功耗、稳定性等一系列技术挑战尚有赖于年轻一代解决。而中国普通高等教育中，还没有“并行计算”这个专业。
说到底，我们需要完善高性能计算的生态环境，才能实现超级计算机产业的健康、可持续发展。“神威·太湖之光”带给中国的惊喜以及中国夺取入选数量上的“头把交椅”的确给了我们在超级计算机领域的自信和勇气，不过，在还没有培育出健康的生态环境之前，我们最好还是不要以“超级计算机第一强国”自居。倒不如踏实做好眼下该做的事，蓄积实力，等待水到渠成的那一天。