Rich in papers, China still less innovative

BEIJING, June 29 (Xinhua) -- I met six U.S. information technology experts in a single week, not even a single one failed to remind me of the prowess of the China-made supercomputer, Tianhe-1A which had occupied the world's fastest in computing for half a year before it was overtaken last week by a Japanese K system.

In the latest edition of the world's top 500 supercomputers, Tianhe-1A and another Chinese brand Nebulae took up two superior slots in top 10, the second and the fourth respectively. Together with another 59 China-made systems, they won Team China the second best, next to the United States, in developing supercomputing technologies.

The international science community often cites how many fastest supercomputers one country is able to create for proving national research superiority. Sounds reasonable. Supercomputer manufacturing requires high-calibre research capability as well as sophisticated technologies and engineering. Supercomputers are used to process big data for, just name a few, finance, aerospace, geophysics, weather and climate forecasts, logistic services and those for defense purposes.

Statistics seemed to tell that China did a fairly good job.

Wait a moment. One vital feature people might neglect is that all the China-made bodies employed American hearts, either Intel or AMD central processing units (CPUs). Years after a leading Chinese CPU manufacturer touted to equip indigenous chips into Chinese supercomputers, most of Chinese CPUs are still mounted to drive low-end laptops and set-top boxes. However, Chinese computing architects did make some gizmos to upgrade system functionality and energy efficiency for those supercomputers.

Another evidence, which is often referred to describe China's advances in research power, is the number of research papers.

Science Watch ranked China the fourth, following the United States, Japan and Germany, in producing 719,971 research papers on Science Citation Index journals from 2000 to 2010. But in terms of average citations per paper, a benchmark for merit of research papers, China was nowhere among the world's top 20, with 5.87 which was a far cry from that of the most innovative countries, Switzerland with 16.62, the United States with 15.77 and the Netherlands with 15.37. What China had scored was close to 6.28 of Brazil and and 5.62 of India.

Even for the lower-cited papers first authored by Chinese researchers, most citations were done by home peers and, in many cases, were self citations.

Statistics also showed one quarter of the total Chinese-first authored research papers in the past decade were in collaboration with international colleagues, mostly American scientists.

The huge scientific research and technological development (R&D) investment in China mainly contributed to the explosive increase in research papers. China exceeded in 2007 the 1.5 percent threshold of R&D share in gross domestic product (GDP), from 0.6 percent in 1996. In most developed economies, the R&D share in GDP is usually around three percent. In absolute calculations, the United States in 2007 invested 369 billion U.S. dollars in research and development, more than that of all Asian countries in aggregate, according to the Organization for Economic Cooperation and Development data.

Research funds in China are used for not only building labs, buying research facilities and recruiting global talents, but also spreading bonus to scientists who are able to publish research papers on peer-reviewed journals, particularly those with high impact factors.

In academic fields like geology and paleoanthropology where the vast land of China and treasure troves allow Chinese scientists a unique access to resources, it is not unusual for home researchers to publish outstanding papers, based on studies on fossils or primitive layers, on even the highest-rated publications, such as Science and Nature.

However, for most research papers, gaining monetary bonus would be priority for their writers. The biggest chunk of all the published papers by Chinese writers were less innovative and might not be based for triggering new thoughts or key technological know-how, which was somewhat testified by patenting.

The U.S. Patent and Trademark Office in 2008 granted 49 percent of patents to U.S. applicants, 22 percent to Japanese, 14 percent to Europeans. Chinese companies and individuals were granted less than one percent of all U.S. patents in the same year. In the most innovative and valuable patents, like inventions, Chinese got only one percent while Americans 30 percent.

That might help explain how less competitive China is in science and technology.

Nonetheless, the wise policy of opening itself to the outside world has made China quite a success, at face value or not. Science in China greatly benefited from the ever easier human and logistic flows in and out.

Take 2007 for example, among all 22,500 who got doctorates in natural sciences and engineering in U.S. institutions, 31 percent originated from China. After getting advanced degrees, some found senior research positions back in China; many of those who chose to stay in the United States played key roles in international partnerships, including churning out research papers. Until those expats change to foreign citizenship, their outputs are calculated for China although they always utilize Western state-of-the-art labs for the most innovative work.

The free flow and exchange now seem to be expanded to bits and bytes. While encouraging its people to chant songs remembering the revolutionary old days, the southwestern city of Chongqing is building a ten-square kilometers IT park where startups and hi-tech multinationals can escape possible internet filtering. The rationale might be mercantile: Far less competitive than India in IT outsourcing services, we need go faster.