New Delhi, India Its portion of a $250 billion semiconductor market is marginal, it has only two fabless startups, its innovation record as measured by the number of patents granted is dismal, it suffers from a dearth of Ph.Ds and high attrition rates, what talent it does have is undisciplined and it let China dominate completely in systems manufacturing.
Yet for Abhi Talwakar, president and CEO of LSI Corporation, India, remains firmly in the driver's seat of its own semiconductor and engineering design destiny.
"India possesses a unique combination of enablers that can greatly elevate its role into the future," he said during his kick-off keynote for this year's VLSI2009 Conference taking place in New Delhi this week.
"There's no reason it can't get on the map [next to Europe and China]," he said, though the US remains clearly in the lead with 48 percent of the market share, based on the number of companies headquartered there. India currently enjoys a meager $7.3 billion of that $250 billion market, or 2.92 percent.
While PCs, consumer and handsets face challenges due to circumstances outside of the industry, he's confident that India will grow, particularly in the areas of automotive, healthcare and energy.
Talwalkar's optimism rests on India's dominance in four criteria, relative to its counterparts in the region, specifically China, Korea and Japan. Those criteria are cost of talent, internal market potential, diversity and scalability and language. While China is comparable on the first three, India wins on language, according to a table he presented.
Not be overlooked are the technology 'anchors' for each country. China has depended on system manufacturing, Korea on memory and Japan on consumer electronics. However, India has depended on software, which, given the increasing role software plays in system design, is of fundamental importance of the country's future, he said.
However, it must be added that according to Sandeep Bhatia, engineering director for Broadcom, India (Bangalore), "there are many different kinds of software," and the software expertise in India to date has been largely focused on financial and IT markets. As such, it will require a shift to adapt to electronic and embedded systems development. Bhatia was speaking as part of a panel later that day on "Accelerating Embedded Systems Development," moderated by TechOnline.
Innovation lacking, not talent
While there is plenty of diverse talent available, Talwalkar said that talent needs more discipline as well as more expertise in the areas of [embedded] software and firmware and VLSI design. "It needs ownership of the entire program," he said.
The theme of India partaking in the entire system development came up a number of times during the conference, and for good reason. As Talwalkar put it, "Only when end-to-end design takes place here will local companies have the ability to influence those designs. However, on a more fundamental level , he pointed out it is only when Indian engineers are exposed to the full system, top down, can they see where the opportunities to innovate lie along the entire design chain.
"India needs 'womb-to-tomb' design, and even marketing," he said. "It also needs more relevant curriculum."
The need to innovate more is borne out by the figures, in terms of patents applied for and patents granted. Leading the pack is the US, with 596,447 and 234,725, respectively. Next comes Israel (6,656 and 1,997), followed by China (4,700 and 1,220) and lastly India, with 2,145 applied for and only 621 granted.
However, being able to innovate also requires a strong background in research, and Talwalkar called for a commitment to that, particularly at the academic institutions, while also recognizing that such a commitment would require stronger links between industry and academia.
Unfortunately, the basic research situation looks to continue, primarily because of a dearth of Ph.D candidates. Speaking during a roundtable discussion on how industry and academia can better cooperate, Professor G.S. Visweswaran of IIT Delhi, made an impassioned plea for better education and more Ph.Ds with a love for basic research.
Professor G.S. Visweswaran of IIT Delhi made an impassioned plea for more Ph.D candidates with a true passion for research, in part at least, so that they don't cheat on their thesis
Too often, he said, candidates get pulled into industry before completing their studies and, more tellingly, too many submit a thesis stuffed with content copied from the Web, indicating a lack of commitment. "We need to get research as a matter of the heart," he said. For that, he also called for more and better-qualified teachers to keep the momentum going and the passion alive.
Fab(-u-)less semiconductor opportunities await
If that core expertise becomes more widely available, along with the confidence to act on it, the opportunities for those graduates are plentiful, according to Talwalkar, particularly in the area of fabless semiconductor design, which now has 1,300 active companies, comprising 20 percent of the overall $250 billion market.
"The foundry industry has been the greatest [game] changer, having drastically lowered the barrier to entry," for semiconductor designers, he said. However, the uneven distribution is India's opportunity. According to Talwalkar, the US has 606, China has 222, Taiwan has 196, Japan has 16, Hong Kong has 16 Singapore has seven, Malaysia has four and India, again, comes in last with two, though he pointed to 28 design services and IP provider houses. "There's the potential for hundreds of fabless companies, but it needs a game changer," he added. Key areas of opportunity for those startups include communications equipment, particularly wireless, as well as data-processing hardware, automotive and industrial.
While developing the required talent with the ability to innovate might solve one problem, another one rises if that talent ends up being too good. "I was surprised to see attrition rates [in India] of 15 to 20 percent," said Talwalkar. That attrition results in a direct cost to the company, not only in terms of lost talent and expertise, but also through the process of rehiring and training.
"It's just not sustainable," he said. "San Jose has attrition rates that are similar and that's a deterrent for large facilities [of 1000 or more]. "
In the end, for India to reach its full potential, Talwalkar made a call to action, not only for industry, academia and government, but also the media, which he said must be especially targeted toward government. "This has to be deliberate in terms of investment, we're talking about a 10- to 15-year investment."