Lessons of China's DSP scandal

masonchung

Beijing gets credit for coming clean on prof's IP theft, but its funding policies helped set the stage for the breach
( ~9 B% Z# m$ l; ~/ u
, G/ X9 x* U' b, `* jBy Mike Clendenin   
- `% P) |, g1 y
" H# y" J+ o7 A( }2 k) ^Courtesy of EE Times
. T0 x5 G& y1 M  M(05/22/2006 0:00 H EST)  

These are dark days for the Chinese chip industry. China confirmed last week that Chen Jin, Microelectronics Department dean at Shanghai's prestigious Jiaotong University, faked the development of a homegrown digital signal processor that China had hailed as a major step forward in its technological capability.
+ p- p( O& d+ c. V# z3 K8 N
1 L& X* M4 l8 T6 [5 h: |
Chen, the 37-year-old project leader for the Hanxin DSP, had been under a cloud of suspicion for more than a year. The university dragged its feet on the investigation, but upon the probe's conclusion, Jiaotong acted swiftly to fire Chen, who allegedly had used intellectual property from his former employer, Freescale. The government and university have ended funding for the DSP effort. Chen is barred from future government-sponsored projects, and some of his peers believe legal action should be taken against him for his misuse of government funds. Some are also calling for an investigation into the conduct of the government panel that certified the Hanxin's authenticity.

) B; h; I% e& K0 [
- R) q# J4 E- D+ F  s, uThere are lessons in this scandal.


. ]1 H/ d) {( ZLesson 1: Review the process and goals of academic projects, and reconsider their tight links to industry. Chen developed the Hanxin in an atmosphere of intense competition for government funds, during a time when officials have been looking for quicker ways to commercialize research. In such an atmosphere, one may promise too much and end up desperate to deliver. Making matters worse, one source said, has been the government's desire to elevate engineers to heroes, to hold up symbols of rapid progress in a profession usually characterized by many small steps.
" {7 B4 p3 w% E

Lesson 2: Don't reinvent the wheel. Why squander government resources on a homegrown DSP, when there are already entrenched offerings in the market? Government money has also subsidized other dubious projects, such as a domestic CPU. Such programs generate headlines but don't make much business sense.
' j. \( l  o' E& c
+ n- K: O9 \4 o- ]
Instead, China should focus on other parts of the food chain, such as software development, or slices of intellectual property that will slot into system-on-chip platforms for popular consumer products. It would be better off lending greater support, with no strings attached, to the private sector to determine which product areas offer the best return for R&D dollars. University work should continue, of course, but should be scrutinized. And the intent should be crafting novel technologies, not Chinese sources for existing ones. That is not innovation.


Jiaotong and the government are to be credited for the high-profile accounting of the Chen investigation. The university pulled no punches, berating Chen for breach of trust, calling his behavior "despicable" and warning other scientists not to travel his path. Investigations into other heavily funded projects are under way, however, and may reveal similar breaches, sources said.


# I0 C" Z9 Z8 f; [9 f9 _As China cleans up the mess, the government must realize that patient encouragement, diligent support and realistic expectations should be the guiding forces for achieving its goal of an innovation society. The country is well on its way, thanks to the work of the many thousands of honest scientists, engineers and educators who are the hope of China's high-tech industry.
% P- h" P# Z6 Z) H- N/ c0 O
: i& J; h# {* u. e6 e3 q[ 本帖最後由 masonchung 於 2007-11-25 12:41 AM 編輯 ]