Published :
Updated :
For all latest news, follow The Financial Express Google News channel.
In getting jobs done better, from a teapot to toilet, every industrial product is the target for advancement through software-running on semiconductors. Furthermore, even the human body is the target to get fitted with silicon chips, whether for knowing health status or augmenting our vital organs to function better. Well, some of us already have a tiny chip underneath the skin of the chest to give kicks to the ailing heart at regular intervals. Yes, the semiconductor has been growing as the cornerstone of further advancement of more or less all products and processes. From incremental upgrades to reinvention, semiconductors will keep redefining the competition space in all conceivable sectors.
Consequently, the semiconductor industry will continue experiencing growth reaching $1 trillion by 2030, from its current size of $600 billion. Hence, new entrants are aggressive in finding an entry route. On the other hand, incumbents are desperate to retain and accentuate their positions-even through massive subsidies and trade restrictions.
Despite the growing volume and indispensable role, every layer of the value chain appears to be extremely monopolised, posing a high entry barrier. On the other hand, geopolitics in forming blue and red supply chains has made it worse. Despite it, powerful incumbents are weakening, and new entrants are growing. Hence, strategising the entry route, and engineering the possibility of creating a snowball or flywheel effect, highly matters.
Due to the large volume and endless growth prospects, some may tell us why we bother about strategising. Just cast your net with whatever resources you readily have. As many less developed countries are blessed with a growing number of electrical and computer engineering graduates and high-wage differentials, jumping into the chip design segment is tempting. But if all the aspiring new entrants enter this segment, will it not get crowded? Furthermore, is there a natural tendency for monopoly, resulting in the winner taking all? Let's draw a lesson from software service export.
We all know that India has succeeded in creating more than 16 million jobs, generating a staggering US$ 133.7 billion in export revenue during 2020-21. Despite having 50 million students in the education systems, as high as 100 universities offering computer science education, and a 20 to 30 per cent wage differential, why has Bangladesh not attained similar success? Like Bangladesh, except the Philippines in voice-based BPO, no other less developing countries have succeeded in having proportionate success. Does it mean that there is a natural tendency of monopoly in exporting technology services? Perhaps, yes. Early entrants enjoy the opportunity of creating economies of scale and scope advantages by leveraging tacit capability and reusable intellectual assets. Hence, late entrants, having a similar education system, student population, and wage differential, face a natural likelihood of failing to repeat the leader's success.
Like the software services in semiconductor design service, India has been an early entrant among the aspiring developing countries. More or less, all well-known semiconductor companies such as Intel, Texas Instruments, and IBM have captive design centres in India. Already more than 20,000 chip designers have been working in these facilities. Furthermore, India has taken a massive program for exposing 85,000+ B.Tech, M.Tech, and Ph.D. students at 120 academic institutions across the country to the chip design area over the next five years. As India will not likely run out of supply, why will semiconductor firms having captive facilities in India and open new ones in other developing countries, like Bangladesh?
Moreover, unlike the software service, clients in the chip design market are limited, as only fabless and integrated device makers are clients. Furthermore, there is intellectual property issue. Hence, despite lucrativeness and apparent success of a few firms creating a few hundred jobs, aspiring developing countries like Bangladesh will likely face steep barriers to scale up-ending up in repeating the disappointment of software service export.
Does it mean that countries like Bangladesh do not have any opportunities in the global semiconductor industry? Perhaps, that is not rational. The challenge for them is to find a suitable entry route and create a snowball effect out of accumulating intellectual assets--building scale and scope effect. The plain vanilla strategy of following the leader for leveraging wage differential will not likely create success. Hence, they need to map out the evolution of the global semiconductor value chain to find answers to some critical questions. For example, how have firms like TSMC, ASML, MediaTek, QUALCOMM, or ARM grown so successful from a humble beginning?
On the other hand, why has Intel been getting weaker? And why have early entrants, many American firms got bankrupt? Similarly, despite having a semiconductor production and research facility established as early as the 1980s and colossal engineering talent, why could India not grow as a semiconductor powerhouse?
Despite significant entry and scale-up barriers, success in the semiconductor industry is transient and keeps migrating. For example, for over 30 years, till 1986, Japan was the largest semiconductor maker. On the other hand, in 1974, through a $10 million contract, Taiwan sowed the seed of the semiconductor industry. Without having often cited success factors like sizeable domestic market, support industry, world-class universities, and lead users, how Taiwan has created such a remarkable success should be the subject of investigation for aspiring countries to draw strategy lessons. On the other hand, after having research fab established in the 1980s, leading users like space and defence programmes, a substantial domestic market, and a large pool of bright engineering graduates, why is India not a visible node in the global value chain? Instead, why is India out there with $30 billion to attract foreign direct investment for kick-starting the semiconductor industry?
As opposed to following others and relying on wage differential, a growing number of engineering graduates, and a few dots of demonstration, aspiring less developed countries should focus on mapping the global space. This exercise is essential to empower them to figure out where they can add value-by making a humble entry and creating a snowball effect. Their mapping exercise may begin with identifying key segments, locating players in each of them, finding how suppliers feed them and defining how those segments are linked within the industry and with the broader economy. It should be followed up by identifying new entrants, market leaders, and dynamics of how firms have been growing and weakening.
Focus on technology, economic, and innovation flow affecting the overall industry structure and competition space will bring further insights. Subsequent investigations should focus on the evolution, gaining market power and disappearance, and acquiring edge by firms and geographies. Upon doing so, this exercise should focus on understanding the unfolding future in terms of disruptions caused by technology, geopolitics, and new entrants-forming the layout of the industry revealing over the next 5 to 10 years.
Once this exercise is done, potential linkage points of the global semiconductor industry with local ecosystem nodes like firms, academic institutions, high-tech park facilities, and the government could be envisioned. Perhaps, such an exercise will form an essential background understanding for developing the strategy of the new entrant for building the semiconductor industry-through making a humble entry and creating a snowball effect. Otherwise, hurriedly planned activities may lead to wasting the opportunity and investment-leading to repeating past disappointments.
Rokonuzzaman, Ph.D is academic and researcher on technology, innovation, and policy