Intel seeks billions for minority stake in Altera business, sources say

[Daniel Nenni]

KEY POINTS

• Intel is looking for an investor that would acquire at least a minority stake worth billions of dollars for its Altera subsidiary, according to sources with knowledge of the matter.
• Intel is seeking a deal that values Altera at around $17 billion, said the people.
• It's possible that Intel would look to find a majority acquirer for the business.

Pat Gelsinger, CEO, of Intel Corporation, testifies during the Senate Commerce, Science, and Transportation hearing on semiconductors titled Developing Next Generation Technology for Innovation, in Russell Senate Office Building on Wednesday, March 23, 2022.
Tom Williams | CQ-Roll Call, Inc. | Getty Images

Intel is looking to sell at least a minority stake in its Altera unit in a transaction that would raise several billion dollars in cash for the struggling chipmaker, according to people familiar with the matter.

Intel is seeking a deal that values Altera at around $17 billion, said the people, who requested anonymity to speak freely about confidential information. Intel purchased Altera for $16.7 billion in 2015.

Following a steep drop in its stock price and extended stretch of market share losses, Intel has been looking to make drastic changes. The company made overtures to a number of private equity and strategic investors this week about Altera, the sources said. Intel has expressed to some of those investors that it would be possible to acquire a majority stake in the business.

A representative for Intel declined to comment. The sale process represents an abrupt change from Intel's prior commentary on Altera. As recently last month, CEO Pat Gelsinger said that Intel's leadership considered the business to be a core part of Intel's future.

Intel has previously said it could look to monetize Altera business through an IPO, possibly as soon as 2026. But the idea of taking strategic or private equity investment would be a marked acceleration of those plans.

Gelsinger and his leadership team have previously said that Intel understands its disadvantaged position and is working aggressively to remedy it. Selling a stake in Altera might allow Intel to more easily pursue its semiconductor fabrication ambitions and assure investors that it has a future as an independent company.

But the sale process also comes as Qualcomm has expressed interest in acquiring its onetime rival, a deal that would face fierce regulatory scrutiny and potentially reshape the semiconductor industry.

Intel stock has dropped 50% this year, as the company has been trounced by Nvidia in artificial intelligence chips and has lost share to Advanced Micro Devices in its core PC and data center market. Shares of Intel were up more than 1% in late-morning trading Friday.

Intel seeks billions for minority stake in Altera business, sources say

This move that would represent a sharp about-face from its earlier public messaging about the larger company's prospects.

www-cnbc-com.cdn.ampproject.org

 

[Daniel Nenni]

Altera was a bad acquisition. Intel needs to sell it off completely, and shed the other distractions to its core business of x86 and manufacturing. I think Altera would do well as an independent company away from Intel bureaucracy and get that start-up vibe going again.

 

[blueone]

For Altera to be successful independently, or otherwise, it probably needs a new CEO and COO. People who are passionate about FPGA technology, and not a former chief people officer and a marketing person. For such a deeply technical product, Altera has a curious leadership team.

 

[Daniel Nenni]

For Altera to be successful independently, or otherwise, it probably needs a new CEO and COO. People who are passionate about FPGA technology, and not a former chief people officer and a marketing person. For such a deeply technical product, Altera has a curious leadership team.

Definitely. Altera needs to go back into start-up mode. I worked with Altera before and after the Intel acquisition which was horrendous. A scrappy company became a corporate bureaucracy and lost sight of what they were doing. It really was hard to watch. Before the acquisition things got done, afterwards there was meeting after meeting and nothing got done.

Now that Xilinx is owned by AMD, Altera has a big opportunity. Using Intel 18A Altera could retake the lead and run with it. Packaging is key for FPGAs and Intel EMBID could be a winner. Altera would have no other competition. Lattice and Achronix have no chance against a re-energized Altera. If I was 10 years younger I would jump aboard Altera in a heartbeat. I worked for an FPGA startup that was acquired by Atmel back in the day. It was a wild ride. The programmable market is an adventure!

Video: Intel EMIB Technology Explained

Intel's multi-die interconnect bridge (EMIB) is an approach to in-package high-density interconnect of heterogeneous chips.

www.intel.com

 

[Xebec]

Altera was a bad acquisition. Intel needs to sell it off completely, and shed the other distractions to its core business of x86 and manufacturing. I think Altera would do well as an independent company away from Intel bureaucracy and get that start-up vibe going again.

This is a good example where an industry insider has a much better view than an outsider. (As an outsider), I always saw Altera's acquisition as a really good thing for Intel - keep feeding the fabs, and push back the eventual barrier (for an IDM) that will be $revenue to justify the ever-increasing cost of fabs.

However, I never considered how much the "Intel corporate culture tax" would reduce.. Altera itself. (It feels like it also took 10 years for the AMD acquisition of ATi to really crystalize into something).

 

[Daniel Nenni]

This is a good example where an industry insider has a much better view than an outsider. (As an outsider), I always saw Altera's acquisition as a really good thing for Intel - keep feeding the fabs, and push back the eventual barrier (for an IDM) that will be $revenue to justify the ever-increasing cost of fabs.

However, I never considered how much the "Intel corporate culture tax" would reduce.. Altera itself. (It feels like it also took 10 years for the AMD acquisition of ATi to really crystalize into something).

I also worked with ATI before they were acquired and yes it was not a smooth acquisition. I remember AMD called the ATI people Red Team and the AMD people Green Team. Not really a team building experience. ATI and Nvidia were fierce competitors but after the AMD purchase not so much.

 

[hist78]

Altera was a bad acquisition. Intel needs to sell it off completely, and shed the other distractions to its core business of x86 and manufacturing. I think Altera would do well as an independent company away from Intel bureaucracy and get that start-up vibe going again.

"Altera was a bad acquisition."

Does Intel have a "good" acquisition, especially for those big acquisitions?

 

[Jozo035]

I see few problems, or disappointments... First would be that everyone kind of expected high-performance x86 core integrated with FPGA fabric trough high bandwidth and low latency link. Maybe even on ringbus or shared cache.

...
Now that Xilinx is owned by AMD, Altera has a big opportunity. Using Intel 18A Altera could retake the lead and run with it. Packaging is key for FPGAs and Intel EMBID could be a winner. Altera would have no other competition. Lattice and Achronix have no chance against a re-energized Altera. If I was 10 years younger I would jump aboard Altera in a heartbeat. I worked for an FPGA startup that was acquired by Atmel back in the day. It was a wild ride. The programmable market is an adventure!
...

This is second point. We expected that Altera will have more resources and they will use that to move to new nodes first. As this was the rule in the past, "first team on more advanced node wins". But both are on 7nm nodes.

Third point is software, which is improving but still remains probably main limiting factor in bigger and more complex devices.

Should software company line Amazon or Google buy FPGA company and put heavy weight in optimization/algorithmic development on the problem? Or will FPGA companies open their toolchains?

 

[tooLongInEDA]

This is just so frustrating.

Yet again (Wind River, Infineon wireless unit, Altera, McAfee, ...), Intel buys a company for a huge amount of money, declares this a critical strategic fit and essential for future product plans and growth, ploughs on for around a decade, destroys value and then disposes of the business. Almost always at a large loss. And then you have bizarre cases like MobilEye which is now half in and half out.

I call this the "hokey cokey strategy".

There is clearly a serious failure of either initial strategy (the acquisition) or execution (post acquisition integration and excution). Or possibly both. It's happened too many times to simply be bad luck.

And curiously, none of the execs responsible for any of this ever get to walk the plank. Perhaps I'm missing something, but the only Intel exec who's been fired in the past couple of decades while all this was going on was Brian Krzanich. And that was for misconduct, rather than inadequate performance. Meanwhile, Lip-Bu Tan, who certainly wasn't responsible for any of this, has gone.

I do agree that this could be a great opportunity for Altera to re-emerge as a leading FPGA company. But in that case, it should be completely independent of Intel. Enough of this hokey cokey nonsense.

 

[nghanayem]

"Altera was a bad acquisition."

Does Intel have a "good" acquisition, especially for those big acquisitions?

If you count IMS as "big", I think that was pretty reasonable. It started as an investment so IMS could afford to develop their multibeam writers in 2009 with intel buying the rest of the shares in 2015. Now it is valued at $4.3B, unfortunately I can't find how much intel paid for it. Maybe it was so small that it was never disclosed? Regardless of the raw value creation, seems like a good investment either way for the technological value of making sure that technology came to market in a timely manner (similar to when Samsung, intel, and TSMC invested in ASML to make sure funding was secure for EUV development).

I guess I would also call the purchase of DEC's semiconductor division a "win". $700M to not spend an eternity in court (and maybe lose/be forced to destroy the IPs) over "hyperthreading" patent litigation, getting like 800 experienced chip designers, and getting a sizable (for the time) 8in fab to boot that would go on to run various specialty technologies for well over a decade. Obviously it wasn't money intel "wanted" to spend. But it seems like intel at least milked a good bit of value out of a measly $700M. Building a new 8in fab in the 90s alone were often hovering around $1B.

And yes, the comedic value of one of intel's best purchases having been one that was forced upon them as a court settlement is not lost on me.

 

[Joseph Byrne]

"Altera was a bad acquisition."

Does Intel have a "good" acquisition, especially for those big acquisitions?

Aviators have a saying: "Any landing you can walk away from is a good landing." McAfee, Mobileye, and now maybe Altera are all deals the company has unwound and been able to walk away from--not unscathed but not with a total loss, either. Among the company's other deals, it's hard to point any with unequivocal positive ROI.

 

[Joseph Byrne]

This is just so frustrating.

Yet again (Wind River, Infineon wireless unit, Altera, McAfee, ...), Intel buys a company for a huge amount of money, declares this a critical strategic fit and essential for future product plans and growth, ploughs on for around a decade, destroys value and then disposes of the business. Almost always at a large loss. And then you have bizarre cases like MobilEye which is now half in and half out.

I call this the "hokey cokey strategy".

There is clearly a serious failure of either initial strategy (the acquisition) or execution (post acquisition integration and excution). Or possibly both. It's happened too many times to simply be bad luck.

And curiously, none of the execs responsible for any of this ever get to walk the plank. Perhaps I'm missing something, but the only Intel exec who's been fired in the past couple of decades while all this was going on was Brian Krzanich. And that was for misconduct, rather than inadequate performance. Meanwhile, Lip-Bu Tan, who certainly wasn't responsible for any of this, has gone.

I do agree that this could be a great opportunity for Altera to re-emerge as a leading FPGA company. But in that case, it should be completely independent of Intel. Enough of this hokey cokey nonsense.

My VC professor used to say bankers never admit to making bad loans--the loans turned bad *after* they made them.

 

[hist78]

Aviators have a saying: "Any landing you can walk away from is a good landing." McAfee, Mobileye, and now maybe Altera are all deals the company has unwound and been able to walk away from--not unscathed but not with a total loss, either. Among the company's other deals, it's hard to point any with unequivocal positive ROI.

If a pilot keeps making heart-stopping landings, even nobody gets hurt, I guess there are two possible outcomes:

1. If this particular pilot is working for a commercial airlines, he/she will be fired or asked to retire. Or, the airlines will face endless lawsuits and FAA investigations.

2. If he/she is a private pilot flying for fun, I think the spouse or kids will try to sabotage the airplane to make it not even be able to leave the hanger. @Daniel Nenni probably has more authority to answer it.

 

[blueone]

For Altera to be successful independently, or otherwise, it probably needs a new CEO and COO. People who are passionate about FPGA technology, and not a former chief people officer and a marketing person. For such a deeply technical product, Altera has a curious leadership team.

Apparently PG is playing musical chairs again. The chief operating officer of Altera is "moving on". Saw his announcement on LinkedIn. Very odd for so soon after the spin-out announcement. It should be interesting to see who the replacement is, assuming Altera/Intel announces it.