Strategies for Semiconductor manufacturing in India
Date: Monday , November 17, 2008
Though semiconductors is now a relatively mature $ 200 billion a year worldwide industry dominated by a few large players, new technologies and applications are still emerging regularly and certain segments show growth of 30 percent per year. India simply cannot ignore this high tech industry that provides the building block for high tech civilian as well as defense hardware, in part because of China’s growing role in it.
In a future trade or military confrontation India’s software skills and revenues there from may prove to be inadequate to compensate for its lack of hardware production muscle.
For the success of its much-publicized Moon Program (“Chandrayaan”) India would need access to semiconductor based sensors, logic and actuators of specialized designs to minimize the weight and improve accuracy of its space vehicles. This was perhaps recognized in the recent takeover of the two-decade old and obsolete Government Semiconductor Complex in Chandigarh and the decision to upgrade it. However the emphasis there still seems to be production of chips currently in vogue (e.g. for smart cards and mixed signal) by importing the latest semiconductor process and design technologies rather than staying abreast via sustained development efforts in - house.
Driven by their obsession to catch up with Japan, their one–time colonial ruler, the South Koreans boldly embarked on manufacture of semiconductor memory as early as the mid 1980s. At that time they had hardly any PhD scientists and quite predictably the whole attempt crashed culminating in the liquidation of Hyundai Semiconductor in the late 1990s.
Taiwan was a relative latecomer to the semiconductor game but since the late ‘90s have became a world-class provider of manufacturing services for the most complex digital processors. Before launching their wafer fabs Taiwan first carried out sustained academic and industrial research over two decades in targeted areas of semiconductors and then attracted back numerous Taiwanese employed in the U.S. semiconductor industry. Since the early 2000 China on the other hand has launched several semiconductor wafer fabs using semiconductor technologies it obtained by arm-twisting foreign MNCs that were lured by China’s cheap labor. China’s wafer fabs today are still far behind those of Taiwan. China’s failure to innovate on their own is however greatly masked by the Chinese governments insistence on use of domestically produced semiconductors—even for outsourced assembly of consumer electronics. Singapore too tried to launch its own wafer fabs but without the academic strength of Taiwan or the outsourcing leverage of China has so far failed to make headway.
With its long history of academic work in theoretical and applied Physics, India too should have been able to create a domestic semiconductor industry with deep expertise along the lines of Japan or Taiwan. But as yet there have not been any serious effort in India to develop a competence driven semiconductor industry, only short bursts urged by a desire to appear not too obsolete.
The success of Taiwan’s strategy should provide confidence that it is never too late to start if an astute and sincere strategy is followed. It is not yet too late to select Semiconductor products and technologies most appropriate for India and then pursue a “no excuses” development program for future leadership in select segments. Semiconductors have a range of applications, not just in computer chips that everyone is familiar with these days but even to generate electricity or efficient light bulbs. If selected and executed carefully, domestic manufacture of relatively immature electronic components used in consumer products can lower costs compared to imports even from China and within 5-years establish India firmly as a semiconductor heavyweight.
Over the last two years there has been a resurgence of interest in semiconductor manufacturing in India—especially around the IT boomtown of Bangalore. Buoyed by their success in exporting software-intensive services, IT professionals in South India, oblivious or undaunted by the tribulations of the Semiconductor Complex in North India (though now controlled by Bangalore-based ISRO), have shown great enthusiasm in launching new domestic semiconductor manufacturing facilities or wafer fabs. Some of these professionals are employed in the electronic/chip design outfits that have recently mushroomed around the IT hotspots in the south. They feel that domestic manufacture of semiconductor chips would be a natural extension to further the returns from their intellectual effort into outsourced chip design or driver software.
December 2004 saw the kick off of the Indian Semiconductor Association (ISA), an advocacy group devoted to promoting start up of semiconductor wafer fabs in India. This was paralleled by the signing of paper agreements between a certain South Korean promoter and the government of an up and coming state in India to launch a billion dollar wafer fab! Not to be outdone by a mere South Korean, NRIs resident in the U.S. descended on India the next winter (Dec ’05) and paraded around those very states promising multi–billion dollar state of the art wafer fabs, thereby managing to arouse yet more irrational euphoria among the local media and state governments, blissfully innocent of even the basics of the hardware business. Despite various fantasy promises of billion dollar investments just around the corner and transfer of state of the art fab know-how from US semiconductor majors, nothing much has come to pass. Savvy semiconductor tycoons in the U.S., including even those of Indian origin have been reluctant to invest in India for semiconductor fabs (though they were very quick to start chip software/ design operations in Bangalore).
It is now widely recognized that the success of the IT industry in India was triggered by the Y2K panic and its subsequent phenomenal growth has been sustained by India’s till now unique ability to provide world class service and customer satisfaction at cut rate prices as well as operate in English. So long as the satellite communication links between Bangalore and the US worked, none of the usual weak links in India (physical isolation from major markets—being blocked off from the US by China and from Europe by the Middle East, governmental intransigence coupled with lack of leverage with foreign companies, absence of venture capital, basic infrastructure and supply lines) hardly mattered. The same dynamic however does not operate for semiconductor manufacturing. The initial investment of several billion dollars needed for a state of the art semiconductor fab established at a green field site like India may not yield any returns for 5 years—by which time the capital-equipment in this fast moving industry may become obsolete! These are not very attractive investment prospects for either the newly minted IT billionaires of India or for that matter investors anywhere.
Advocates of creating a viable and world class semiconductor industry in India, including the ISA, should do well to lobby the Government to immediately empanel a “blue–ribbon” committee consisting of content experts and businesspeople to develop a 5 year plan for semiconductors and then commit to set up at least 3 competing semiconductor development centers around the country. A total investment of up to $ 300 million will be worthwhile, as it would finally create the deep pool of expertise that will give India an honest claim at technical excellence and then enable new industries and quality jobs worth many billions.
The author is Chief Technology Officer, APSTL
He can be contacted at firstname.lastname@example.org