‘With right support, we can design and test microchips in the country’

Right now the global superpowers are fighting a war, but not a traditional one. It is a boardroom war to prevent rivals from getting their hands on the most cutting edge technology in the world — microchips, also known as semiconductors, integrated circuits or silicon chips. 

These tiny pieces of chips are the work-horse for all modern technologies from artificial intelligence, machine learning, cloud computing to IoT and advanced medical devices. Most things you use in your day-to-day activities likely have semiconductor chips in them. 

Anything that has any technological "smartness" to it is a product of semiconductors. Not just the obvious techy things, but also things like your refrigerator, microwave and digital clock — all have semiconductor chips in them. 

Today, it is estimated that we use 130 to 140 chips per person on the planet on average. This number is going to be more than a thousand by the end of this decade. As the market grows, everyone wants a slice of it. But where is Bangladesh in the game? The answer is a blunt "nowhere yet". 

However, it is not all doom and gloom. 

Four organisations, Ulkasemi, Neural Semiconductor, Prime Silicon and Toton Electronics, are working on designing chips in Bangladesh. 

Nadim Chowdhury, an MIT graduate and assistant professor at Buet, wants to bridge the industry-academia gap to prepare engineers for the future chip industry and build an intellectual grip on semiconductor technology. 

Nadim grew up in Khulna, Chattogram and Dhaka. After completing his HSC in 2006 from Chittagong College, he started his undergraduate studies at Buet in 2007. 

He graduated top of his class in 2012 and joined Buet as a lecturer. In 2015, he went to MIT to pursue his PhD and second master's degree from the department of EECS (Electrical Engineering and Computer Science). 

From his undergrad thesis to PhD, Nadim studied advanced semiconductor technologies. After completing his PhD, he returned to Bangladesh, turning down job offers in the US. 

Nadim Chowdhury, assistant professor, Buet

There are two reasons behind Nadim's return to Bangladesh after his time at MIT. One was his realisation that Bangladeshi students were missing out on 'true' research experiences, which he aims to provide to the best of his capacity and experience gained in the west.

"Ever since my undergraduate years, I have been into the research of semiconductor devices. In our country, most research is conducted through simulations, involving the development of computational frameworks to assess device performance. This entails solving a series of partial differential equations. We utilise tools like Matlab, Python and other online simulation resources," Nadim explained.

His amazement was palpable when he saw real transistors, witnessing their physical characteristics, applications, and manufacturing processes. He described the experience as "truly astonishing". 

Transistors are minuscule, beyond the resolution of standard microscopes, necessitating Scanning Electron Microscopes (SEM) to observe them in microchips, given their scale falls below the wavelength of visible light.

Nadim's return to Bangladesh came shortly after his successful PhD defence. He also won the best PhD thesis award at Microsystems Technology Laboratories, MIT in spring 2022. 

The other reason for his homecoming was that he aspired to reshape academia's contribution to the country's economy. "I want to catalyse change within our country's semiconductor industry. In Bangladesh, you will see university faculties rarely initiate startups and scarcely establish industry connections. Even if such connections exist, they are generally at a consultative level, at best."

Taking a glance at some of the world's most prominent companies reveals that many originated as university spin-offs. For instance, Moderna traces back to MIT, Dropbox too emerged from MIT, and Google has its roots in Stanford. "Unfortunately, this entrepreneurial culture remains scarce within our own universities. My goal is to foster this culture and drive change from Buet," he added.

What can be done

Manufacturing microchips is a daunting task for a country like Bangladesh right now. But Nadim argues that is our end goal. Maybe we will build our chips in 2050. But to take our country to that level, we need to build the ecosystem, build the knowledge, build the deft manpower. Our baby steps may start with chip design, testing and packaging.

For Bangladesh, he emphasised the importance of chip design, given the global shortage of chip designers and engineers. He saw potential in training engineers for chip design, which could eventually lead to the creation of a thriving industry. 

He drew parallels to Taiwan's success in the semiconductor sector and highlighted the need for support from the government and collaboration with industry experts to nurture this field in Bangladesh.

As the use of microchips only increases, it will be a huge market with different subsectors. For example, we are not using chips in our clothes yet. There is a team at MIT working on smart fabric. You will be wearing your regular shirt, but that will monitor your temperature, pulse and all other details to suggest any action to be taken to prevent bad things from happening.  

"Smart fabric is coming up very soon. Since we have a strong position in the garment industry, we have a chance to grab this market before others. So, apart from manufacturing garments, we need to put more stress on R&D in the textile industry," said Nadim. 

"Since we cannot manufacture chips yet, for starters, we can start with chip designing. We already have a few design firms that have started working with this." 

It works this way: for example if a company needs a chip, it will be needing circuit designers to design it. In Bangladesh, we have yet to complete design chips for clients. We can do partial designs only. There are expert Bangladeshi engineers who can do that, but they don't stay in Bangladesh now. 

"We have to create new chip engineers and retain our engineers with high facilities like Taiwan did," he said.  

Moreover, the world now has a massive shortage of chip engineers. According to a recent New York Times report, there is a huge chip engineer shortage globally. 

The report says that the US semiconductor industry could face a shortage of about 70,000 to 90,000 workers over the next few years. It presented a shortfall of about 300,000 engineers and 90,000 skilled technicians in the US alone by 2030.

"We produce on average 20,000 electrical and computer engineering graduates every year. If we can train only 1,000 of them for chip designing, we will have 10,000 chip designers in 10 years and that will translate to half a billion dollar industry. We are already late. If we don't do it right now, we will regret it," Nadim said. 

Some people think that the silicon wafer is expensive. But it is not that expensive. Even the wafer business is not that lucrative nor complex. The magic happens with what you put on the wafer. 

Taiwan is the only country in the world that can make 2 nanometre chips. They have an edge because of their geopolitical location; Taiwan's strategic location in East Asia has allowed it to forge close ties with major technology markets such as China, Japan, and South Korea. But their success can be credited mostly to their engineering wonders. 

"Now what we need is cumulative knowledge. See it this way, if you want to solve a problem from class 10, you have to grow through class one, two and gradually the 10th class. You might skip one or two levels somewhere but it is not possible that you will be starting from the top. The chip knowledge is like that. We have to start building our own knowledge," he further explained.

Chip designing is not like python and coding that you can learn online. There are no YouTube tutorials, no one shares this knowledge. Even Taiwan has established strong intellectual property protection laws and regulations, which have contributed to safeguarding their intellectual hold. 

In our academic curriculum, there are very few elements for semiconductors. But it should be studied even broadly. We need to update our courses, and hire exceptional faculty members to teach and research semiconductors.  

We need faculties who worked with semiconductors (completed their masters and PhD abroad), and did some real work. People who worked with companies like Intel, Analog Device Inc, GlobalFoundries, and who did some advanced study in the field, also who work with technology development for chips. Nadim also suggests arranging training sessions for our existing engineers. 

Additionally, we can make chip testing labs. "We were offered to do chip testing by a US company. But we could not take the offer because we do not have a testing lab," he said.  

There is a myth that a chip testing lab is so expensive that we cannot bear the expenses. But from a country's point of view, it's not that expensive. "If we can invest $10-20 million, we can have a very well-functioning chip testing lab," he opined.

What Nadim has achieved so far

With his return to Bangladesh, Nadim took significant steps towards realising his vision. He secured funding for projects like the nanolab and chip design training, paving the path for the semiconductor industry to take root in the country. His collaboration with MIT and other institutions allowed Bangladeshi engineers to contribute to cutting-edge research.

Nadim admires figures like Morris Chang, who built Taiwan's semiconductor industry through a combination of practical knowledge gained abroad and strategic application back home. 

"Chang is a Taiwanese-American engineer. He did not learn chip engineering in Taiwan, he was an MIT graduate. He worked for a quarter of a century at Texas Instruments and then joined the Taiwan research body that spun off TSMC (Taiwan Semiconductor Manufacturing Company Limited).  He learned hands-on knowledge in the US and then applied it to Taiwan and developed the knowledge to an unreachable height. Today, TSMC customers are big companies like Apple, Nvidia, and Qualcomm," said Nadim. 

Nadim dreams of replicating this model in Bangladesh by fostering an environment where experts could focus solely on advancing semiconductor technology. 

His efforts extended beyond academia. He engaged in lobbying for local chip design companies on the international stage, leveraging his specialist status to garner attention and opportunities for these companies. 

He has a plan to establish a startup in the semiconductor industry in Bangladesh which stemmed from the observation that faculty members in the country rarely established startups.

As Nadim transitioned to his role as an assistant professor, he contemplated dedicating more time to research, balancing it with teaching responsibilities. He is optimistic about the future of chip design in Bangladesh. He believes that the country's talented students, when provided with the necessary resources and government support, could excel in the field of semiconductor research and development.