Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/apollo-acquires-49-of-intels-fab34.20361/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Apollo acquires 49% of Intel's Fab34

I did not consider that. What ballpark is that moneywise? Depreciation on a new fab?


1. If we assume the fab 34 land, building, and clean room cost about $4 billion.
2. Intel said it has spent $18.4 billion on the Fab 34 already. That means Intel probably spent about $14 billion on the equipment and installation.
3. Intel uses a 8-year depreciation schedule (vs TSMC's 5-year) for equipment. Using the above $14 billion equipment number will come out $1.75 billion yearly depreciation expense.
4. If we don't include the building and clean room depreciation cost, Intel Ireland Fab 34 will need to come out at least $1.75 billion depreciation cost a year for next 8 years.

You can change above number to whatever more appropriate or accurate figures.
 
Last edited:
No one notices that Apollo buys 49% stake of Fab 34 with 11B, that means the total investment of the fab should be 22B and Intel already invested 18.4B so the deal actually will lock 7.2B for Intel to invest in other fabs?
 
1. If we assume the fab 34 land, building, and clean room cost about $4 billion.
2. Intel said it has spent $18.4 billion on the Fab 34 already. That means Intel probably spent about $14 billion on the equipment and installation.
3. Intel uses a 8-year depreciation schedule (vs TSMC's 5-year) for equipment. Using the above $14 billion equipment number will come out $1.75 billion yearly depreciation expense.
4. If we don't include the building and clean room depreciation cost, Intel Ireland Fab 34 will need to come out at least $1.75 billion depreciation cost a year for next 8 years.

You can change above number to whatever more appropriate or accurate figures.
First of all, fab 34 construction cost has never been just 4B, according to EIS, it planned to spend 8B on the construction.
 
No one notices that Apollo buys 49% stake of Fab 34 with 11B, that means the total investment of the fab should be 22B and Intel already invested 18.4B so the deal actually will lock 7.2B for Intel to invest in other fabs?
I think the rest of the cash will be headed for Germany early next year.
 
No one notices that Apollo buys 49% stake of Fab 34 with 11B, that means the total investment of the fab should be 22B and Intel already invested 18.4B so the deal actually will lock 7.2B for Intel to invest in other fabs?

In the private equity investment or investment in general, the total value of the business is not necessary a simple calculation of investment amount and share percentage.

For example, in this Intel-Apollo deal, Apollo didn't get a piece of ownership for the land, building, clean room, and equipments. Intel still owns and operates Ireland fab 34. What Apollo will get is a percentage of the profit generated from the Fab 34 chips manufacturing. In some complex investment structure, certain investors may get more or less true rewards than the share percentage they own.
 
Last edited:
Some more details about the Intel-Apollo Ireland Fab 34 deal from Intel's 8-K filing with SEC.


"Item 8.01 Other Events.

In connection with the execution of the Purchase and Sale Agreement and Limited Liability Company Agreement, the parties will enter into certain ancillary agreements which generally provide Intel or Intel Ireland, as applicable, with rights and obligations with respect to the construction, commissioning, operation, management, maintenance, utilization and purchase of wafers produced in Fab 34 consistent with their ordinary course practices with respect to such matters but subject to agreed minimum wafer fab performance standards and minimum wafer purchase obligations by Intel. Intel Ireland will retain full ownership of Fab 34 and any and all intellectual property rights arising from, relating to or in connection with the activities of and services to JV Company. Specifically, these ancillary agreements provide that:

• JV Company will acquire from Intel Ireland rights to cause and direct the operation of Intel Ireland’s Fab 34 facility, allowing JV Company to direct the manufacture of wafers from Fab 34 for the benefit of JV Company;

• JV Company will engage Intel Ireland to operate and maintain Fab 34, with Intel Ireland agreeing to minimum performance standards including with respect to production, availability, and yield, and to provide administrative services to JV Company;

• Intel Ireland agrees with JV Company and Co-Investor to complete the construction and commission of Fab 34, including the achievement of certain timing and capacity milestones, with substantial completion of Fab 34 targeted for June 2026;

• JV Company and Intel Ireland will allocate risk of loss and set out insurance matters relating to Fab 34 whereby JV Company will bear the risk of certain portions of certain categories of losses;

• Intel agrees to purchase wafers from JV Company for itself and/or for marketing and sale to third-party customers, with Intel agreeing to minimum volume commitments for its demand for wafers, preferential loading of Fab 34 relative to other Intel fabs running the same process and tooling, and pricing of wafers at manufacturing cost plus an agreed upon margin, and the payment of liquidated damages in various circumstances if Intel fails to meet its minimum volume commitments; and

• Intel provides a parent guarantee with respect to the obligations of Intel Ireland.

Under the various ancillary agreements, Intel or Intel Ireland may be required to pay liquidated damages or termination payments under a variety of underperformance, breach, termination or other scenarios in such amounts as to enable JV Company to distribute in respect of Co-Investor's units an amount up to 140% of the Purchase Price."


Note: The bold typeface are modified by me but the text content are original.

Source: https://www.intc.com/filings-report...0000050863-24-000103/0000050863-24-000103.pdf
 
Last edited:
Some more details about the Intel-Apollo Ireland Fab 34 deal from Intel's 8-K filing with SEC.


"Item 8.01 Other Events.

In connection with the execution of the Purchase and Sale Agreement and Limited Liability Company Agreement, the parties will enter into certain ancillary agreements which generally provide Intel or Intel Ireland, as applicable, with rights and obligations with respect to the construction, commissioning, operation, management, maintenance, utilization and purchase of wafers produced in Fab 34 consistent with their ordinary course practices with respect to such matters but subject to agreed minimum wafer fab performance standards and minimum wafer purchase obligations by Intel. Intel Ireland will retain full ownership of Fab 34 and any and all intellectual property rights arising from, relating to or in connection with the activities of and services to JV Company. Specifically, these ancillary agreements provide that:

• JV Company will acquire from Intel Ireland rights to cause and direct the operation of Intel Ireland’s Fab 34 facility, allowing JV Company to direct the manufacture of wafers from Fab 34 for the benefit of JV Company;

• JV Company will engage Intel Ireland to operate and maintain Fab 34, with Intel Ireland agreeing to minimum performance standards including with respect to production, availability, and yield, and to provide administrative services to JV Company;

• Intel Ireland agrees with JV Company and Co-Investor to complete the construction and commission of Fab 34, including the achievement of certain timing and capacity milestones, with substantial completion of Fab 34 targeted for June 2026;

• JV Company and Intel Ireland will allocate risk of loss and set out insurance matters relating to Fab 34 whereby JV Company will bear the risk of certain portions of certain categories of losses;

• Intel agrees to purchase wafers from JV Company for itself and/or for marketing and sale to third-party customers, with Intel agreeing to minimum volume commitments for its demand for wafers, preferential loading of Fab 34 relative to other Intel fabs running the same process and tooling, and pricing of wafers at manufacturing cost plus an agreed upon margin, and the payment of liquidated damages in various circumstances if Intel fails to meet its minimum volume commitments; and

• Intel provides a parent guarantee with respect to the obligations of Intel Ireland.

Under the various ancillary agreements, Intel or Intel Ireland may be required to pay liquidated damages or termination payments under a variety of underperformance, breach, termination or other scenarios in such amounts as to enable JV Company to distribute in respect of Co-Investor's units an amount up to 140% of the Purchase Price."


Note: Bold text are done by me.

Source: https://www.intc.com/filings-report...0000050863-24-000103/0000050863-24-000103.pdf

So where does that leave Intel?

In your summation what does "an agreed upon margin" come in at?

If the margins come under pressure who gets squeezed first , Intel or their suppliers?
 
So where does that leave Intel?

In your summation what does "an agreed upon margin" come in at?

If the margins come under pressure who gets squeezed first , Intel or their suppliers?

All text in the italic are from Intel's 8-K filing. I didn't summarize anything.

If you ask me my feelings about what this document has revealed, I have only one word: "sad".

It's very sad that Intel needs to resort to this kind of contract to get $11 billion cash infusion.
 
Simply put, due to the CHIPs Act and other subsidies they are building too much capacity, which they can't fund through cashflow as they historically did.

The fact that this is all happening at the same as their core business is being taken away from them means it is game over for Intel as most of us here have known them.

Intel will probably still be around in 10 years time but they will be an also-ran, nothing more.
 
Simply put, due to the CHIPs Act and other subsidies they are building too much capacity, which they can't fund through cashflow as they historically did.

The fact that this is all happening at the same as their core business is being taken away from them means it is game over for Intel as most of us here have known them.

Intel will probably still be around in 10 years time but they will be an also-ran, nothing more.
Will totally depend on their fabs and their product teams being competitive a reason to use TSMC is to not loose buisness and than bring it back inhouse as much as possible
 
So where does that leave Intel?

In your summation what does "an agreed upon margin" come in at?

If the margins come under pressure who gets squeezed first , Intel or their suppliers?

Let's try to figure out what "an agreed upon margin" means. Because the Intel-Apollo contract is confidential, I can only use public information to estimate the "trend" or "theme" of this deal. We can discuss the method and number or you can do your own analysis but please don't jump on me for my calculations (impossible to be accurate).

First, what's the investment return a Private Equity asking for? According to Google AI search:

"The annual return rate for private equity (PE) can vary from year to year, but some studies have shown that it can outperform other investments over the long term:

• Cambridge Associates' U.S. Private Equity Index: PE had an average annual return of 14.65% from 2000 to 2021, which is higher than the 11.53% average annual return for venture capital (VC) over the same period. However, VC outperformed PE from 2010 to 2020, with an average annual return of 15.15%.

• Cliffwater: PE returned 11% annually over a 23-year period, which is higher than the 6.2% annualized return for public stocks. Over a 22-year period ending in 2022, state pension allocations to PE produced an 11.4% annualized return, which is 5.6% higher than the 5.8% annualized return for public stocks.

• S&P Listed Private Equity: PE returned 36.07% in 2023, -23.62% in 2022, and 55.20% in 2021. "



That means the Apollo "may" receive 15% ~ %20 annual return. Again, this is my own guess and please feel free to plug in your own number.

It means Intel's payment obligation to Apollo is at least $1.65 billion annually ($11 billion * 15%). Is it expensive? Let's do some comparison.

Again, I use Google AI search:

"As of June 7, 2024, the rates for some corporate bonds were:
• Moody's Seasoned Aaa Corporate Bond Yield: 5.19%, lower than the long term average of 6.46%
• US Corporate BBB Effective Yield: 5.67%, higher than the long term average of 5.27%
• US Corporate AAA Effective Yield: 4.89%, higher than the long term average of 4.06
%"

Also, in May TSMC completed a NT$11.5 billion unsecured green bond sale:

"TSMC said it will use the proceeds from the upcoming green bond sale to finance spending on green architecture projects and other environmental protection plans.

According to TSMC, the unsecured NT$11.5 billion green bond sale, which its board of directors approved last week, will consist of two tranches: a five-year NT$4.9 billion issuance with a coupon rate of
1.94 percent, and a 10-year NT$6.6 billion issuance with a coupon rate of 2.1 percent."

Source: https://focustaiwan.tw/business/202405130009

Another case is Micron Technology.

"On January 12, 2024 Issuer Micron Technology issued international bonds (US595112CD31) with the coupon rate of 5.3% in the amount of USD 1000 mln maturing in 2031. The issues were sold at the price of 99.93% at par. The bookrunners of the placement were Morgan Stanley, Credit Agricole CIB, Bank of Nova Scotia, CIBC, HSBC, Mitsubishi UFJ Financial Group, TD Securities, Truist Bank, Wells Fargo."


Source: https://cbonds.com/news/2663615/

Just for the annual payment alone, this Apollo deal is not cheap to Intel especially if we compare it with Intel's peers.

Private equity fund investment typically has an exit in about 5 to 6 years. It makes me wonder what's in Intel's mind. An IPO for Intel Foundry or some kinds of spinoff?
 
Last edited:
Will totally depend on their fabs and their product teams being competitive a reason to use TSMC is to not loose buisness and than bring it back inhouse as much as possible

Unfortunately, the Intel-Brookfield and Intel-Apollo financing deals make the Intel product division harder to compete. Because the wafer volume commitment and profit/return guaranteed by Intel, Intel product division will be forced to use the Intel fabs involved in the contracts.

Intel and Pat Gelsinger had stated that in the IDM 2.0 and Smart Capital strategies, Intel Product division is free to choose between internal fab and external foundries in order to compete. In reality, it's gone.
 
Last edited:
You build some fab shells to collect US and Euro Chips Act money, while stealthily going fab lite. It has to be stealthy, so stealthy. The handouts and private financing wouldn't be there if you were not planning to operate the shells. It wouldn't be the first time an IDM built a shell in Europe and never operated there. But in the US, this is new.
 
Will totally depend on their fabs and their product teams being competitive a reason to use TSMC is to not loose buisness and than bring it back inhouse as much as possible
I guess my underlying point is that have no realistic chance of catching up with TSMC as a foundry business, either by process or practice. Let alone create any clear blue space.

They missed the smartphone but the cloud saved them. Now they've missed the switch to GPU but this time nothing is going to save them.

Intel without monopoly profits is not the Intel we (or they themselves) know.
 
Unfortunately, the Intel-Brookfield and Intel-Apollo financing deals make the Intel product division harder to compete. Because the wafer volume commitment and profit/return guaranteed by Intel, Intel product division will be forced to use the Intel fabs involved in the contracts.

Intel and Pat Gelsinger had stated that in the IDM 2.0 and Smart Capital strategies, Intel Product division is free to choose between internal fab and external foundries in order to compete. In reality, it's gone.

Yes, the flexibility is gone, but the intention has always been that Products will use outside for fringes, so that IF doesn't have to cover everything.

if after 5N4Y and IF is still uncompetitive to run the core of Intel Product's wafers, then this 4 year experiment would have been a failure.

The stated goal has always been that Intel will regain both product and process leadership (or at least , competitiveness.)
 
Let's try to figure out what "an agreed upon margin" means. Because the Intel-Apollo contract is confidential, I can only use public information to estimate the "trend" or "theme" of this deal. We can discuss the method and number or you can do your own analysis but please don't jump on me for my calculations (impossible to be accurate).

First, what's the investment return a Private Equity asking for? According to Google AI search:

"The annual return rate for private equity (PE) can vary from year to year, but some studies have shown that it can outperform other investments over the long term:

• Cambridge Associates' U.S. Private Equity Index: PE had an average annual return of 14.65% from 2000 to 2021, which is higher than the 11.53% average annual return for venture capital (VC) over the same period. However, VC outperformed PE from 2010 to 2020, with an average annual return of 15.15%.

• Cliffwater: PE returned 11% annually over a 23-year period, which is higher than the 6.2% annualized return for public stocks. Over a 22-year period ending in 2022, state pension allocations to PE produced an 11.4% annualized return, which is 5.6% higher than the 5.8% annualized return for public stocks.

• S&P Listed Private Equity: PE returned 36.07% in 2023, -23.62% in 2022, and 55.20% in 2021. "



That means the Apollo "may" receive 15% ~ %20 annual return. Again, this is my own guess and please feel free to plug in your own number.

It means Intel's payment obligation to Apollo is at least $1.65 billion annually ($11 billion * 15%). Is it expensive? Let's do some comparison.

Again, I use Google AI search:

"As of June 7, 2024, the rates for some corporate bonds were:
• Moody's Seasoned Aaa Corporate Bond Yield: 5.19%, lower than the long term average of 6.46%
• US Corporate BBB Effective Yield: 5.67%, higher than the long term average of 5.27%
• US Corporate AAA Effective Yield: 4.89%, higher than the long term average of 4.06
%"

Also, in May TSMC completed a NT$11.5 billion unsecured green bond sale:

"TSMC said it will use the proceeds from the upcoming green bond sale to finance spending on green architecture projects and other environmental protection plans.

According to TSMC, the unsecured NT$11.5 billion green bond sale, which its board of directors approved last week, will consist of two tranches: a five-year NT$4.9 billion issuance with a coupon rate of
1.94 percent, and a 10-year NT$6.6 billion issuance with a coupon rate of 2.1 percent."

Source: https://focustaiwan.tw/business/202405130009

Another case is Micron Technology.

"On January 12, 2024 Issuer Micron Technology issued international bonds (US595112CD31) with the coupon rate of 5.3% in the amount of USD 1000 mln maturing in 2031. The issues were sold at the price of 99.93% at par. The bookrunners of the placement were Morgan Stanley, Credit Agricole CIB, Bank of Nova Scotia, CIBC, HSBC, Mitsubishi UFJ Financial Group, TD Securities, Truist Bank, Wells Fargo."


Source: https://cbonds.com/news/2663615/

Just for the annual payment alone, this Apollo deal is not cheap to Intel especially if we compare it with Intel's peers.

Private equity fund investment typically has an exit in about 5 to 6 years. It makes me wonder what's in Intel's mind. An IPO for Intel Foundry or some kinds of spinoff?
A PE firm will likely use leverage in their deals to boost their return. Some leverage buyout deals could have 10x leverage. So they can charge a rather low yield while still generating high return.

Investing in such a deal with Intel is buying a loan. Banks invest in loans all the time. BB rating loans could have a yield of 6-7% now. With certain level of leverage, it's possible to achieve 15-20% ROE in good times.
 
Yes, the flexibility is gone, but the intention has always been that Products will use outside for fringes, so that IF doesn't have to cover everything.

if after 5N4Y and IF is still uncompetitive to run the core of Intel Product's wafers, then this 4 year experiment would have been a failure.

The stated goal has always been that Intel will regain both product and process leadership (or at least , competitiveness.)

With the promised profit and wafer volume given Apollo, Intel product division will be forced to use Intel Ireland fab 34 and stay on Intel 3/4 even if fab 34 is not a competitive option.

Even worst, Intel needs to send its precious cash to Apollo when the market demand is in a down cycle in order to make up any profit promised.

The whole thing is less to do 5N4Y. It's more to do with Intel's obsolete IDM business model and Intel/Pat Gelsinger's egos.

I am amazed that Intel still can't understand a true foundry partner (or partners in general) is NOT someone you call only when you are in trouble or when you have some leftover food after a party (and that "someone" wasn't invited).

Let me quote Nvidia CEO Jensen Huang recent comments about TSMC's potential price hikes. I don't think any Intel customers ever bothered to say such words.

""Raising prices, I think, is consistent with the value that they deliver. And so I'm very happy to see them succeed," Huang said." ~ Nvidia CEO Jensen Huang

Source: Thread 'Nvidia is making a ton of money. Now, main supplier TSMC may charge Jensen Huang more for chips' https://semiwiki.com/forum/index.ph...may-charge-jensen-huang-more-for-chips.20396/

Without cherishing the external foundry relationships, Intel will be treated as a third class customer. How can Intel compete?

Does Intel still think all its competitors (most of them are fabless) are just lucky? After 30+ years, does Intel still think the foundry model is a fluke?

Does Intel still think to build a $20 billion fab first and then customers will come automatically?

Does Intel know more Chips Act subsidies went to non-Intel companies? Does Intel understand that the US government is preparing for a new US semiconductor industry without the Intel used to be?
 
Last edited:
Back
Top