2011-3-3 11:21:00

Reported by Zhao Yongxin

The People’s Daily  (March 2, 2011)

At 2 o’clock PM of the 23rd day of the last month of the lunar calendar in 2010, Hu Leijun, the technical director of Inspur, went out the R&D center which locates in No.2 of Shangdi Information Road, Haidian District, Beijing with his colleagues, and headed home to celebrate the Chinese New Year.

Though every one of them was extremely tired, they could hardly hide their triumph: after hard efforts and struggles of more than 1000 days and nights, they overcame a perilous peak in the field of server at last: the first formal prototype of the 32-way high-end fault-tolerant server (computer) in China was put through and worked well, and all performances could meet the planned targets.

This made Inspur become one of the small numbers of manufacturers mastering the technology of high-end fault-tolerant server after those international major companies such as IBM and HP. This not only marked that Chinese server had entered the world first camp, but also indicated that the situation of high-end servers in Chinese key economic fields such economy and telecom being monopolized by foreign brands would be broken.

Server stores and handles 80% of the data and information on the network, and is known as the "nerve center" of network. The high-end fault-tolerant server with strong information processing power and high reliability can be used in banks, telecommunications and civil aviations, etc, and its importance is self-evident. Its market value is also very considerable: the price will reach several hundred thousand and even more than million Yuan easily, which is the most lucrative part of the server market.

However, it’s unluckily that until now, the market of high-end fault-tolerant server of our country has always being monopolized by the foreign manufacturers such as IBM and Hp.

In the eleventh five-year plan, the country has made a strategic decision to research and produce the high-end fault-tolerant server independently. After a selection, the science and technology department had determined Inspur and Huawei to undertake the R&D of the "863" high-end fault-tolerant computer project in January 2008. The state invested 260 million Yuan; Shandong government invested 150 million Yuan, Inspur invested 337.5 million Yuan, and it was 747.5 million Yuan in total. It is reported that this is by far the largest investment in R&D project of our country's information industry.

For this project, Inspur chose 100 excellent technicians and set up 10 R&D bureaus. There were two roads to choose: one was to cooperate with the foreign manufacturers which master the core technology. It was safe and could save time and effort, but would always be enslaved; another road was to do it yourself and take all the risks. They finally chose the latter.

The average age of R&D group is less than 30. They made many breakthroughs in a lot of core technologies:

——proposed “Large twin blades extensible and tightly-coupling server system architecture” creatively, and mastered the international advanced technology which can complete the 64-way IP Peering only by 1 level leapfrog;

  ——Developed independently the first global high-speed interconnecting chip in server and solved the problem of data transfer among CPUs. Its data exchange and transmission rate had reached the international advanced level;

——Kept pace with the R&D of hardware and developed their own high-end server operating system. So they got rid of the foreign control in this level...

On July 28, 2010, Inspur successfully developed the first key technology identification prototype of 32-way high-end fault-tolerant server; in August of the same year, this prototype has a on-line testing in China Construction Bank, and its key technology had reached the expected level.

It is reported that the 32-way high-end fault-tolerant server developed by Inspur is hopeful to realize its volume production in this year. Speaking of its application prospect, the senior vice president of Inspur, Wang Endong is not so optimistic.

He analyses that for a long time, key business systems such as banks and telecoms have been relying on the foreign products, and most part of them are locked in the foreign closed systems and it makes it difficult for them to transplant. “Besides, for the domestic self-owned brands are at disadvantageous market position, most of them can hardly compatible with the hardware and software of foreign brands. That’s also a problem.”

“Creation is a complete chain. Besides supporting technology and product R&D, the county should do more on supporting the new product promotion. At the same time, it is needed for the domestic consumers to be brave and be the first one to eat the crab, that is to purchase and use domestic brands positively”, said Wang Endong.