The Vedanta Semiconductor Plant Has Gone with the Wind

Not so long ago, Vedanta Chairman Anil Agarwal announced in a series of tweets that in a joint venture with Taiwanese multinational electronics contract manufacturer Foxconn, it had signed a Memorandum of Understanding (MoU) with the Gujarat government to set up a semiconductor and display manufacturing unit. Agarwal said that the landmark investment of INR 1.54 lakh crore would help make India’s Atmanirbhar Silicon Valley a reality. 

However, as per recent reports, Vedanta is struggling to get financial backing to facilitate the establishment of these units. Additionally, ET reported that Agarwal’s representatives met with large financiers from the Middle East, Singapore, and the US over the last three months to secure funding commitments for the project but to no avail. 

This follows Agarwal’s admission that the project’s funding would come from debt and that the business had no immediate plans to sell shares. But, there is still no clarity as to why Vedanta will look for foreign backing when several reports populated this year that the conglomerate is in talks with banks to raise a debt of $2.5 billion to $3 billion for its semiconductor and display manufacturing units. 

Arun Mampazhy, a veteran semiconductor analyst, told AIM that he speculates that debt-raising plans from Indian banks didn’t work out for Vedanta, which forced them to look for external funding. Along with overseas investment as well, multiple factors could be charted to its failure:

• The technical inexpertise of Vedanta’s technical knowledge partner, Foxconn.
• Foxconn unable to keep on many of the promises it has made in recent times.
• Foxconn still considered to be soft on China. 

Foxconn’s shaky efforts

The terms of the partnership say that Vedanta will hold 60% of equity to Foxconn’s 40% for the semiconductor fabrication site. Foxconn was brought on board to be a technical partner in the project. But, one only wonders why, since the “Electronics Manufacturing Services” company is not known to have any technical know-how of running a 28nm fab. While it has been a subcontractor in terms of managing assembling units for the company’s products like Apple’s iPhones, Microsoft’s Xbox and Sony’s Playstations, it has no experience in running a foundry. 

Moreover, reports also emerged that Ministry for Electronics & Information Technology (MeitY) asked Vedanta and Foxconn technology to provide details on how they plan to get the required expertise to run a plant at the scale that the government envisions or to get a third partner that could demonstrate technical capability. 

Foxconn has also been notorious for announcing projects that are now, as Mampazhy says, either “still in the making” or intentionally keeping it under wraps. For example, in 2016, it acquired the Japan-based Sharp, which, along with having Sharp’s OLED screen panels, also owns a 200mm wafer fab running the 130nm process nodes. While several reports suggested that it will slowly move from 200mm fab to 300mm fab (28nm technology) in a facility in Zhuhai, there is no update on the current status of Zhuhai fab at the moment. 

In May this year, Foxconn acquired an indirect stake in the Malaysian semiconductor manufacturer SilTerra to build a 28nm fabrication plant in Malaysia. However, the existing infrastructure is only for 200mm wafers, which is not enough to run a 28nm fab. Additionally, there were also speculations that it is in talks with established fabs like TSMC or UMC for its 28nm technology or, as Mampazhy notes, with an Intermediate Research Organisation (IRO) like Interuniversity Microelectronics Centre (IMEC) Belgium. 

According to a report, the Indian government has also signed an agreement with IMEC to provide the 28nm in exchange for a royalty. 

IRO serves as a link between academic institutions and businesses, but they are independent and receive funding from a variety of public and/or private sources. As a result, they play a crucial role as mediators, fostering innovations, licensing intellectual property, starting spinoffs, sharing know-how, pooling R&D resources, enhancing workforce capabilities, networking, and labour exchange operations.

But, as Mampazhy notes, “What is obtained as technology is typically not a production ready one but a development stage one which may take two–three more years of further maturing to get to the yield levels required for high volume manufacturing.” And, regarding its plans with TSMC or UMC, it is not known yet how they will materialise when it comes to India. Therefore, while the Vedanta-Foxconn venture initially showed high promise, the current situation portrays a rather uncertain future.

Towards IRO Localisation? 

When AIM had a conversation with K Krishna Moorthy, CEO, and president of IESA, he emphasised the importance of building a parallel R&D for the new fabs to be profitable in the next five to seven years and lead the way for the next-generation technology. Mampazhy, on the other hand, disagrees with this and argues that the need of the hour is to leverage the potential offered by the existing IRO like ITRI in Taiwan, IMEC in Belgium, and several in the United States, like SUNY University, since it is not easy to come up with a business model for IRO. Hence, right now, the priority is to get the mature node fabs in production—by all means.