Rapidus pursues US$690 million loan to achieve 2nm chip production by 2027

[Daniel Nenni]

Credit: AFP

Rapidus, aiming to begin mass production of 2-nanometer chips in Japan by 2027, is seeking significant financial support for its trial production and eventual mass production.

The company is in talks with a trio of major Japanese banks—MUFG Bank, Sumitomo Mitsui Banking Corporation, and Mizuho Bank alongside Development Bank of Japan (DBJ) for a loan of approximately JPY 100 billion (US$690 million). Rapidus is also seeking additional investment from its existing backers, including Toyota.

Filling in the chips funding gaps​

According to sources cited by Kyodo News, the four banks Rapidus is currently negotiating with have all previously provided substantial financing to support the Japan-based Kioxia. Their efforts helped stabilize the finances of the NAND Flash manufacturer to remain competitive in the market.

Rapidus was established in 2022 with JPY 7.3 billion in joint funding from eight Japanese companies, including Toyota, Sony, NTT, NEC, SoftBank, Denso, Kioxia, and MUFG Bank. Additionally, the Japanese government has pledged subsidies totaling JPY 920 billion over three years, which has enabled Rapidus to construct its first factory in Chitose, Hokkaido.

However, Rapidus estimates that trial production of the 2-nanometer chips by 2025 will require JPY 2 trillion, with an additional JPY 3 trillion needed for mass production by 2027. This forecast reveals a funding gap of over JPY 1 trillion for the trial production phase alone, prompting Rapidus to seek loans and additional investment from Toyota and other stakeholders.

Despite these efforts, challenges remain in successfully mass-producing 2-nanometer chips, particularly in terms of technology and securing a sufficient customer base. This has led to cautious attitudes among corporations and financial institutions regarding financing or further investment in Rapidus.

There is no guarantee that Rapidus will secure the loans from the four banks. The Japanese government, under Prime Minister Fumio Kishida, is considering legal revisions that would allow the government to provide guarantees to assist Rapidus in securing private financing.

However, with Kishida set to step down by late September or early October 2024, it remains uncertain whether the government will expedite the legal changes to support Rapidus before the next prime minister takes office.

Rapidus pursues US$690 million loan to achieve 2nm chip production by 2027

Rapidus, aiming to begin mass production of 2-nanometer chips in Japan by 2027, is seeking significant financial support for its trial production and eventual mass production.

www.digitimes.com

 

[MKWVentures]

Credit: AFP

Rapidus, aiming to begin mass production of 2-nanometer chips in Japan by 2027, is seeking significant financial support for its trial production and eventual mass production.

The company is in talks with a trio of major Japanese banks—MUFG Bank, Sumitomo Mitsui Banking Corporation, and Mizuho Bank alongside Development Bank of Japan (DBJ) for a loan of approximately JPY 100 billion (US$690 million). Rapidus is also seeking additional investment from its existing backers, including Toyota.

Filling in the chips funding gaps​

According to sources cited by Kyodo News, the four banks Rapidus is currently negotiating with have all previously provided substantial financing to support the Japan-based Kioxia. Their efforts helped stabilize the finances of the NAND Flash manufacturer to remain competitive in the market.

Rapidus was established in 2022 with JPY 7.3 billion in joint funding from eight Japanese companies, including Toyota, Sony, NTT, NEC, SoftBank, Denso, Kioxia, and MUFG Bank. Additionally, the Japanese government has pledged subsidies totaling JPY 920 billion over three years, which has enabled Rapidus to construct its first factory in Chitose, Hokkaido.

However, Rapidus estimates that trial production of the 2-nanometer chips by 2025 will require JPY 2 trillion, with an additional JPY 3 trillion needed for mass production by 2027. This forecast reveals a funding gap of over JPY 1 trillion for the trial production phase alone, prompting Rapidus to seek loans and additional investment from Toyota and other stakeholders.

Despite these efforts, challenges remain in successfully mass-producing 2-nanometer chips, particularly in terms of technology and securing a sufficient customer base. This has led to cautious attitudes among corporations and financial institutions regarding financing or further investment in Rapidus.

There is no guarantee that Rapidus will secure the loans from the four banks. The Japanese government, under Prime Minister Fumio Kishida, is considering legal revisions that would allow the government to provide guarantees to assist Rapidus in securing private financing.

However, with Kishida set to step down by late September or early October 2024, it remains uncertain whether the government will expedite the legal changes to support Rapidus before the next prime minister takes office.

Rapidus pursues US$690 million loan to achieve 2nm chip production by 2027

Rapidus, aiming to begin mass production of 2-nanometer chips in Japan by 2027, is seeking significant financial support for its trial production and eventual mass production.

www.digitimes.com

Rapidus said they will have pilot line in 2025. This needs to be the priority for people like me who have trouble believing a company with no fab manufacturing experience can start at 2nm. If they are still looking for funding, this would be a problem.

 

[Mooredaddy]

The head scratcher that is rapidus continue. Really do not see the business case for this venture.

 

[Daniel Nenni]

I think national pride is a big part of it. I just don't think IBM is a good partner for semiconductor high volume manufacturing. Also, building a competitive foundry ecosystem is an expensive and daunting task. One of the reasons the semiconductor IP segment is growing much faster than EDA (look at the Synopsys 2024 numbers thus far) is the foundries building ecosystems. Intel, Samsung, and now Rapidus. It is an IP bubble for sure but for the greater good, absolutely. On the flip side TSMC gets all IP free and first.

Samsung and Intel paid hundreds of millions of dollars to port IP and no customers showed up at 3nm and 2nm. An IP port is millions of dollars of NRE so without follow-on customers IP companies don't make money. Fool me once shame on you. Fool me twice shame on me.

 

[Daniel Nenni]

IBM lay-offs at China R&D teams send shock waves across local tech community

Chinese employees at US computing giant IBM, once considered a cradle of mainland engineers, have expressed dismay over a terse conference call with the company's American executives on Monday, as the firm known by the nickname Big Blue cuts hundreds of jobs at two local research labs. The...

www.yahoo.com

 

[Jert]

National pride, of course. Agree.

Rapidus's main differentiation is cycle time. They target high mix low volume. Of course resulting cost/ASP per die will be much higher.

I don't think they are fantasizing to compete directly with TSMC (or even Samsung).

 

[BruceA]

National pride, of course. Agree.

Rapidus's main differentiation is cycle time. They target high mix low volume. Of course resulting cost/ASP per die will be much higher.

I don't think they are fantasizing to compete directly with TSMC (or even Samsung).

High mix low volume means low yield, questionable quality and not real

 

[MKWVentures]

National pride, of course. Agree.

Rapidus's main differentiation is cycle time. They target high mix low volume. Of course resulting cost/ASP per die will be much higher.

I don't think they are fantasizing to compete directly with TSMC (or even Samsung).

Like Intel, their goal should be to pass HH Grace in wafer foundry revenue by 2027.
We can talk about cycle time when they run a wafer. Pilot line 2025, correct? so tools installs are well underway, correct?

 

[samwilde]

I will be eating popcorn while reading about this train wreck of an enterprise. That is for sure.
Why not just have TSMC build a line to make chips for Playstation consoles for Sony or something like that. At least that would make sense.

This is just a prestige project for the Japanese government which will turn into a white elephant.

 

[osnium]

I will be eating popcorn while reading about this train wreck of an enterprise. That is for sure.
Why not just have TSMC build a line to make chips for Playstation consoles for Sony or something like that. At least that would make sense.

This is just a prestige project for the Japanese government which will turn into a white elephant.

If Intel shuts down the fabs and sells off the equipment for 5 cents on the dollar, maybe they can transfer technology and tools to Rapidus

 

[Mooredaddy]

I will be eating popcorn while reading about this train wreck of an enterprise. That is for sure.
Why not just have TSMC build a line to make chips for Playstation consoles for Sony or something like that. At least that would make sense.

This is just a prestige project for the Japanese government which will turn into a white elephant.

It will certainly be a disaster in my opinion. 2nm was just an arbitrary goal they set out for without a business case or any realistic way to get there. The whole idea of a boutique leading edge fab is insane in its self. The whole point is SCALE to make the economics work in a fab, especially so at the leading edge to ramp and yield.