Intel Terminates Tower Semiconductor Acquisition

[Daniel Nenni]

Hello Daniel,

Intel just announced that it has mutually agreed with Tower Semiconductor to terminate Intel’s proposed acquisition of Tower due to the inability to obtain in a timely manner the regulatory approvals required under the merger agreement. You can find the press release here.

Intel CEO Pat Gelsinger commented on today’s news:

“Our foundry efforts are critical to unlocking the full potential of IDM 2.0, and we continue to drive forward on all facets of our strategy. We are executing well on our roadmap to regain transistor performance and power performance leadership by 2025, building momentum with customers and the broader ecosystem and investing to deliver the geographically diverse and resilient manufacturing footprint the world needs. Our respect for Tower has only grown through this process, and we will continue to look for opportunities to work together in the future.”

Intel is building a differentiated customer value proposition as the world’s first open system foundry, which will advance its goal of being the second-largest global external foundry by the end of the decade.

Since its launch in March 2021, IFS has gained significant traction with customers and partners, including:

· Agreeing with Synopsys to develop a portfolio of IP on Intel 3 and Intel 18A

· Winning the U.S. government’s RAMP-C award, with RAMP-C customers in design engagement on Intel 18A including Boeing, Northrop Grumman, IBM, Microsoft, and Nvidia

· Landing a multigeneration agreement with Arm to enable chip designers to build low-power compute SoCs on 18A

· Signing a strategic partnership with MediaTek to use IFS’ advanced process technologies

· Adding a leading cloud, edge and datacenter solutions provider as a leading-edge customer for Intel 3

· Growing revenue more than 300% y/y in Q2’23

Please let me know if you have any questions.

Sophie Won (she/her)
Financial Communications

 

[hist78]

It is a setback for Intel and one more reason/excuse for US to impose more sanctions against PRC.

 

[cheesehead]

Guess we are going to see some exciting announcements in IFS soon...

Random thought: couldn't Intel just go ahead without Chinas approval? It is Probably an important Part of the acquisition contract. They are american though and Tower is Israeli as far as I know. Such a move would likely cause further escalation of the on-going economic cold war, but well, does that really matter at this point?

 

[Daniel Nenni]

I don't think it is a problem for Intel. It would have been nice to stuff some foundry business into IFS but Intel needs to get the leading edge process foundry business in order to make a mark. Tower Semiconductor was a "nice to have" not a "must have", my opinion.

 

[Mooredaddy]

I don't think it is a problem for Intel. It would have been nice to stuff some foundry business into IFS but Intel needs to get the leading edge process foundry business in order to make a mark. Tower Semiconductor was a "nice to have" not a "must have", my opinion.

I thought you point over the last year was that the institutional knowledge of Tower was what was the point and value add of this acquisition? Of course Tower has nothing to add in terms of leading edge but isn't this a setback for Intel to not get this help in getting the core structure and operating procedures from Tower to help IFS?

 

[Daniel Nenni]

I thought you point over the last year was that the institutional knowledge of Tower was what was the point and value add of this acquisition? Of course Tower has nothing to add in terms of leading edge but isn't this a setback for Intel to not get this help in getting the core structure and operating procedures from Tower to help IFS?

I think the Tower acquisition was a strong point that Intel is in the foundry business for the long term and that point has been made. I also think having the Tower sales channel would have been good for Intel mature technologies but not so much the leading edge (18A). And as I said stuffing some revenue in IFS would have been nice.

China is taking the blame for this one so Intel is good. I do believe, however, Intel would not have made the Tower offer today under the current economic/foundry market conditions, my opinion.

 

[Blue Devil]

I think the Tower acquisition was a strong point that Intel is in the foundry business for the long term and that point has been made. I also think having the Tower sales channel would have been good for Intel mature technologies but not so much the leading edge (18A). And as I said stuffing some revenue in IFS would have been nice.

China is taking the blame for this one so Intel is good. I do believe, however, Intel would not have made the Tower offer today under the current economic/foundry market conditions, my opinion.

It sounds like you believe losing the Tower deal is a net positive for Intel? I assume circling back to a deal with GF is still off the table?

 

[Daniel Nenni]

It sounds like you believe losing the Tower deal is a net positive for Intel? I assume circling back to a deal with GF is still off the table?

It cost Intel $300M so it was not without pain but yes I think Intel will do better without Tower (or GF) for now. Wall Street wants big customers and that will require 18A and below, that should be the focus. Notice the IP deal with Synopsys is for Intel 3 and 18A. Nothing above that. Focus focus focus........

 

[MKWVentures]

I think this is a big hit to IFS. Yes the focus is on the future technologies. .... but Intel is not a top 15 foundry today. Most of their IFS revenue comes from non-wafer sales (Packaging and mask writing). The revenue from wafer sales is less than 100M/quarter. They need mature technology foundry to build on IF they want to be successful.

Side note: why did Intel change the definition of IFS revenue? Q2 2022 IFS revenue was reported as 122M in Q2 2022. Now Q2 2022 revenue is reported as 57M? and the current revenue is a 307% increase.... but is actually similar to revenue reported for last 2 years?

 

[MKWVentures]

I don't think it is a problem for Intel. It would have been nice to stuff some foundry business into IFS but Intel needs to get the leading edge process foundry business in order to make a mark. Tower Semiconductor was a "nice to have" not a "must have", my opinion.

@Daniel Nenni When do you think Intel will become a top 6 wafer foundry in the world in terms of revenue? 2025? 2027?

When Samsung shows foundry revenue, does that include Samsung processors? or just sales to other companies?

 

[Daniel Nenni]

I think this is a big hit to IFS. Yes the focus is on the future technologies. .... but Intel is not a top 15 foundry today. Most of their IFS revenue comes from non-wafer sales (Packaging and mask writing). The revenue from wafer sales is less than 100M/quarter. They need mature technology foundry to build on IF they want to be successful.

Side note: why did Intel change the definition of IFS revenue? Q2 2022 IFS revenue was reported as 122M in Q2 2022. Now Q2 2022 revenue is reported as 57M? and the current revenue is a 307% increase.... but is actually similar to revenue reported for last 2 years?

Why do you think Intel did not extend the agreement? They could have pushed it out another year?

 

[delong.height]

I don't think it is a problem for Intel. It would have been nice to stuff some foundry business into IFS but Intel needs to get the leading edge process foundry business in order to make a mark. Tower Semiconductor was a "nice to have" not a "must have", my opinion.

I do also think that tower semi is a good add-on if the deal can be done a lot earlier, like last year for example. Now the foundation has been built. Even if the deal gets to passed later this year or next year, there won't be good synergy between IFS and tower because IFS is now entering growth phase with more and more leading-edge customers to be announced later this year or next year. It's ideal for them to cut it now before too late. Tower won't help much at this point in time, since Intel had nearly 2.5 years to find the talent from its internal team and external foundrys or ecosystem.

 

[delong.height]

@Daniel Nenni When do you think Intel will become a top 6 wafer foundry in the world in terms of revenue? 2025? 2027?

When Samsung shows foundry revenue, does that include Samsung processors? or just sales to other companies?

Intel is already the second player if consider its internal demand as fabless clients. Intel has a lot more capacity at Intel 16 that will get more customers over time. I think for Intel 7, Intel will eventually restructure the fab so they can be used for Intel 16. Because Intel 7 isn't competitive to TSMC N7 when it comes to cost while Intel 16 can do so. Also it's a waste of resources trying to develop another process tech in between Intel 16 and Intel 3. So at the end, we are probably talking about a company that has more capacity than Global Foundries at 16, and Intel has a lot more technical strength compared to Global Foundries.

I also remember Intel said previously that it plans to make Intel 16 to be able to transition to Intel 3 as smoothly as possible. So that is also an awesome point to look at.

 

[cliff]

I assume circling back to a deal with GF is still off the table?

A merge between Intel and GF cannot be allowed.

So at the end, we are probably talking about a company that has more capacity than Global Foundries at 16, and Intel has a lot more technical strength compared to Global Foundries.

GF gets to focus on two local/IoT/edge/ASIC/inference lens abundant processes: 22 fdsoi and 14 finfet, and not deal with stacking dies. GF strategically made the right moves. They need to deliver in this sweet spot that they carved for themselves.

Intel's fab needs to compete against the relentless fab and packaging company that their relentless competitor (AMD/Xilinx) is using.

 

[delong.height]

GF gets to focus on two local/IoT/edge/ASIC/inference lens abundant processes: 22 fdsoi and 14 finfet, and not deal with stacking dies. GF strategically made the right moves. They need to deliver in this sweet spot that they carved for themselves.

That is also the direction Intel 16 is going for

 

[Mooredaddy]

I don't think using internal revenue for IFS would be an accurate representation of their foundry business at all. I personally find it extremely unlikely Intel will ever seriously threaten TSMC but more going to eat GF's lunch and parts of SMIC and Samsung.

 

[MKWVentures]

Why do you think Intel did not extend the agreement? They could have pushed it out another year?

They can now start from scratch with a new deal if they wish. They can also restart negotiations with Global or other foundries. Even before the trade war, China was a pain to deal with on merger negotiations.... so maybe they saw minimal hope.

 

[MKWVentures]

I don't think using internal revenue for IFS would be an accurate representation of their foundry business at all. I personally find it extremely unlikely Intel will ever seriously threaten TSMC but more going to eat GF's lunch and parts of SMIC and Samsung.

Intel needs to focus on overtaking Grace and then SMIC in revenue. Their goal is to provide a 3rd leading edge company to TSMC and Samsung. But this is a lot like when IBM wanted to provide foundry services and may have the same result.

 

[hist78]

They can now start from scratch with a new deal if they wish. They can also restart negotiations with Global or other foundries. Even before the trade war, China was a pain to deal with on merger negotiations.... so maybe they saw minimal hope.

There are so many changes are happening at Intel on the product, technology, organization, operation, spending, and business models. Intel needs to cut down the uncertainty as much as possible, including Tower Semi acquisition.

 

[delong.height]

Intel needs to focus on overtaking Grace and then SMIC in revenue. Their goal is to provide a 3rd leading edge company to TSMC and Samsung. But this is a lot like when IBM wanted to provide foundry services and may have the same result.

No, Intel's goal is to be 2nd, just behind TSMC, and given their roadmap, they could potentially do so if they execute. They can pass all other foundries very easily because Intel 16, Intel 3 and Intel 18A are all leading edge. They should be able to have a higher revenue because of that.