India’s Tata Group is working on entering the semiconductor industry. Per a Reuters article, the Indian conglomerate is negotiating a $300 million investment deal for a domestic outsourced assembly and test plant.
It is a certain bit of interesting news. My biggest video ever was about India and its efforts in the semiconductor industry. If you want to watch that, it is below.
So we are going to take a look at the Tata Group and their work in the semiconductor industry.
What is Tata Group?
Tata Group is one of India's biggest multinational business groups. Founded over a century ago, the company is a sprawling conglomerate. They has expanded into a variety of different businesses over the years.
The company's subsidiaries include the world's second biggest manufacturer/distributor of tea (Tata Consumer Products), a $180 billion IT company (Tata Consultancy Services), one of India's biggest steelmakers, and one of its leading automakers - Tata Motors.
Background
Tata has previously said that it wants to do something with the semiconductor industry. Back in August 2021, the company's Chairman - and I am going to butcher this guy's name ... Natarajan Chandrasekaran - spoke at the annual general meeting of the Indian Merchants Chamber of Commerce and Industry.
His remarks centered on the global supply chain being so heavily reliant on China. And he asserted that this is changing, with firms bringing more of that in-house.
Then he said this:
At the Tata group, we have already pivoted into a number of new businesses like electronics manufacturing, 5G network equipment as well as semiconductors, in all probability
These remarks were not recorded and attendees must have been reciting from memory or translating. I have seen his remarks also be quoted in this way.
At the group, we have already set up a business to seize the promise of high-tech manufacturing of electronics, precision manufacturing, assembly and testing, and semiconductors in the medium term
What is OSAT?
Semiconductor manufacturing can be very broadly categorized into two general stages: The front and back-end.
The front end is what most people think about when they think about semiconductor microfabrication. TSMC, Intel, and Samsung Foundry with their billion dollar fabs. What those fabs are largely doing is front end wafer fabrication. They are making the wafers that would turn into chips.
But how do you get from a pizza-sized printed wafer to an actual chip that you plug into your Asus motherboard? Someone has to cut the wafer into chips, encase them within ceramic or plastic so to make it easier to handle, and test them to ensure quality.
These steps are called "assembly" (I have also seen it referred to as “packaging”) and "test". Together make up what we call the "back end".
Workers have to handle special materials or operate a machine. The work is mundane and repetitive but at the same time requires dexterity and close attention. Furthermore, productivity is relatively low. Thus historically, back-end processes were characterized as being relatively low-value compared to the front end.
India's Manufacturing Problem
For one thing, I think that it is a good thing that Tata has actually done manufacturing before. But it does seem like the country has been moving away from manufacturing sectors in recent years and that will have an impact on any attempt to build up a domestic semiconductor industry.
India's economy has grown very fast over the past few decades. The interesting thing about that growth is that much of it has come from services. This differs from East Asia, which have generally achieved their economic success through advanced manufacturing.
For instance, India's most valuable companies are Reliance Industries (a petrochemical company with a telecom subsidiary, Jio), Tata Consultancy, and Infosys.
In 2020, the value-add share of India's economy derived from manufacturing was about 13%. This is down from previous years but even back then that percentage was 15%. And even before the pandemic, the trend had been on the downside.
By comparison, the value-add share of China's economy derived from manufacturing in 2020 was about 27%.
Thoughts
OSAT is a rough industry to enter. If the company’s building a $300 million factory, then they are not going to do it just for themselves. They want to be able to sell those services to other companies like Intel and the likes. And the Reuters article seems to agree:
Potential clients of Tata's OSAT business include companies such as Intel (INTC.O), Advanced Micro Devices (AMD) (AMD.O), and STMicroelectronics (STM.BN).
However, the OSAT industry is right now pretty competitive. As of 2019, 50% of OSAT capacity is located in China and Taiwan. 80% are in Asia. The top 8 companies capture 76% of the industry’s revenue. It is a pretty cut-throat business.
An OSAT facility is a fine first step, but it should not be construed as being an intermediary step into front-end work. Technology transfer is much harder for the front end than the back end. And there are examples of many countries that have succeeded in back-end but flopping at the front-end, leaving them stuck in
Furthermore, there remain the substantial institutional and business barriers that keep the semiconductor industry from making inroads. The lack of substantial capital, high-level talents (particularly in semiconductor manufacturing), and lack of reliable water/power. All of these needs to be addressed before foreign companies will yield to the government’s approaches and bring in their technologies and skills.