Micron's US$150b global expansion plan aims to capture demand for memory

Claudia Tan HS

Published Thu, Oct 21, 2021 · 08:18 AM

EVEN if the pandemic eases and demand for the chips used inside cars and work-from-home gears revert to pre-Covid levels, the memory and storage space will continue to grow exponentially, according to Manish Bhatia, executive vice-president of global operations at semiconductor giant Micron Technology.

Micron had on Wednesday (Oct 20) announced its intent to invest more than US$150 billion globally over the next decade to boost memory manufacturing as well as research and development.

Memory and storage are a growing portion of the global semiconductor industry, and today represent around 30 per cent of the semiconductor market. Secular growth drivers such as 5G and artificial intelligence (AI) are expected to expand usage of memory and storage in areas like automotive and various user devices.

The road maps for the automotive industry, for instance, has shown the need for more advanced memory and storage solutions.

"It really is a data centre on wheels, you're talking about a terabyte of storage, you're talking about 100 gigabytes of DRAM - that's 10 times what you have in your PC - it's a server's worth of DRAM... in your car in order to give you an autonomous experience," Bhatia told The Business Times in an interview.

DRAM or dynamic random access memory is a memory chip that stores data used in computers, workstations and servers.

Micron's investment also comes amid an unprecedented chip shortage that has disrupted many industries, spurring chip makers globally to ramp up capital investments.

Intel Corp had in September said that it plans to build new chip-making facilities in Europe valued at up to US$95 billion. Others such as Taiwan Semiconductor Manufacturing, Samsung Electronics and GlobalFoundries are also racing to boost their supply chains.

"The supply chain shortage is the result of a very steep spike in demand as the world recovers from the pandemic and certain parts of the semiconductor industry just weren't really prepared for that," said Bhatia, adding that it is therefore crucial to plan ahead given that time is required to set up semiconductor factories.

"Micron has done a good job in terms of planning for our capacity growth to be in line with the long-term demand trends that we see," said Bhatia.

Governments too have been doubling down on the semiconductor industry amid the chip shortage or desire to move their economies up the value chain, said Bhatia.

Micron is engaged with governments around the world, including the United States as it evaluates the possibility of fab expansion there.

Memory manufacturing costs in the US is about 35 per cent to 45 per cent higher than in lower-cost places with established semiconductor ecosystems, according to Micron.

Bhatia noted that the semiconductor manufacturing ecosystem in Asia had grown significantly thanks to strong government support. The US, on the other hand, had fallen behind in this aspect.

"So we are engaged with officials in Washington to try and close the cost gap... in terms of building construction costs, infrastructure costs or labour and other input costs," he said, adding that these are necessary areas to improve on to make large-scale manufacturing a reality there.

Micron noted that funding to support new manufacturing capacity and a refundable investment tax credit are critical to the potential expansion in the US.

In Singapore, Micron will be looking into new areas that will complement the local ecosystem. This includes NAND technology, assembly and testing and potentially DRAM technology.

NAND memory is used in devices including smartphones and personal computers, as well as in data centres.

The prospect of introducing DRAM technology here could bring some new elements to Singapore's already rich semiconductor ecosystem that does not exist today, said Bhatia.

Meanwhile, Big Data and AI will continue to shape the manufacturing industry.

Bhatia said these trends revolve around how industry players can use AI and Big Data to enhance control and improve quality output.

Some US$1 billion will also go towards investing in sustainable operations in the next 5 to 7 years.

Resources may be poured into investing in equipment or other software that will help improve emissions, boost recycling efforts or tap renewable energy.