A joint Micron-Western Digital document argues that funding from the US CHIPS and Science Act’s $52 billion pot should be partly used to set up the Memory Coalition of Excellence Recommendations (MCOE) for the National Semiconductor Technology Center.
The Act’s aim is to help strengthen domestic memory (DRAM and NAND) fabrication stateside.
This document is written by five Micron people and five Western Digital people, including president of technology Siva Sivaram.
WD, through its Kioxia joint venture, has no fabs in the US – only in Japan. It is hard to see why the US government should fund domestic R&D to help give the WD-Kioxia’s JV access to new technology for its Japanese fabs.
But if WD were to get closer to Micron – such as sourcing NAND chips from Micron, or even building a joint fab in the US – then that would be a different matter. As far as we can see, the only incentive for it to be involved in this report is if it has an intent to have a closer relationship with Micron.
We wondered if the Elliott Management involvement with WD is a factor here?
Blocks & Files asked WD about this and the reply from its VP corporate communications was: “We appreciate your interest. We don’t have any further comment to provide. Everything we have to say on this matter is in the white paper. Feel free to use that as your source.”
We asked some analysts what their opinions were on why WD is involved with Micron in the MCOE and what the potential outcomes are for WD’s NAND business? Also, on why Micron benefits from WD’s involvement with the white paper. Only one could be quoted directly.
Jim Handy
Jim Handy from Objective Analysis gave us his take on what each party would get out of this.
WDC: The company would not only receive US government grants to finance more R&D and accelerate development, but it would also collaborate with Micron, which should reap synergies. It would need to take measures to ensure that any proprietary Kioxia technology would not fall into Micron’s hands, and vice versa.
Micron: Similar advantage to WDC – more R&D dollars and some synergies. Whatever technology Micron gets would largely be produced in Singapore.
Kioxia: If WDC brings home a new technology, it would probably have to share it with Kioxia, by nature of their JV agreement. While this may not be an absolutely US-centric outcome, the technology would still be out of the hands of Samsung, SK hynix, and YMTC, whose combined NAND market share is 56 percent.
USA: For a modest investment the country would help strengthen two US companies, WDC and Micron, against foreign competitors, even though most of the factory jobs related to these are outside the US. The CHIPS and Science Act isn’t only about manufacturing facilities – it’s mainly about bolstering the competitiveness of US chipmakers in the world market. The MCOE would be a viable way to help in that effort.
Handy added: “Something that always fascinates me is that governments like to sponsor fabs, yet there are several other important steps to chipmaking. Any country that wants to become autonomous in semiconductors would need to do package & test as well, but governments don’t seem to worry about that. The entire supply chain has become very global. Aspirations for autonomy would be extraordinarily challenging to achieve.”
Analyst number 2
Another analyst who spoke on the condition of anonymity said there is a lot of good and bad feeling between Micron and WD management based on years of reorgs and movement between the companies. Sanjay (Mehrotra) and Siva (Sivaram) worked together in the past.
He said that WD and Micron are both memory companies headquartered in the US but they do nearly all manufacturing offshore. So they seem like US memory companies and there are no others. Possibly they partnered in order to get free money for R&D as the only US memory companies available.
If the money is needed to support memory R&D then it helps provide local support. Back in 1987 Sematech was set up as a not-for-profit consortium in the USA to carry out memory technology R&D. He said this was not a successful activity and is not a supporter of government consortiums in general.
His take-away from the MCOE partnership is that if the only US memory companies want free money for R&D then more power to them. But Sematech history suggests it will be a failure.
The money involved comes from the $11 billion NSTC R&D grants and here is a breakdown:
- 11B over five years
- Including National Semiconductor Technology Center (NSTC), National Advanced Packaging Manufacturing Program, and other R&D and workforce development programs;
- FY2022 = $5 billion
- $2B for NSTC
- $2.5B for advanced packaging
- $500M for other R&D programs
- FY2023 = $2B, FY2024 = $1.3B, FY2025 = $1.1B, FY2026 = $1.6B
That fair amount of cash.
Analyst number 3
A financial analyst who also did not want to be directly quoted said that the WD-Micron partnership could be perceived as a move closer by the companies, but also reflective of the perceived long-term risk of the development of memory capabilities in China. That means YMTC.
He also found it interesting that this comes along as WD has remained consistent in its comments that it is engaged in exploring strategic alternatives, including a possible separation of the company that involves discussions with “outside parties.”
Comment
More information about the MCOE may come from Western Digital’s forthcoming investor meetings and quarterly earnings call. Ditto Micron. We have asked MIcron about all this but it hasn’t so far been able to reply.
In general Blocks & Files is, perhaps, over eager to see signals of WD’s break-up and a realignment with Micron in this MCOE partnership affair.
Jim Handy’s words may well be more realistic: “It’s mainly about bolstering the competitiveness of US chipmakers in the world market. The MCOE would be a viable way to help in that effort.”
