The First Big Move at Intel

Lip-Bu Tan has hit the ground running.

Intel’s newly appointed CEO as of March 18 has just made his first major move at the beleaguered semiconductor giant.

Lip-Bu Tan’s First Day on the Job, March 18 | Source: Intel

Announced early this morning, Intel has agreed to sell 51% of its Programmable Solutions Group to private equity firm Silver Lake. As a reminder, Intel’s Programmable Solutions Group is basically Altera – a semiconductor company that Intel acquired back in the summer of 2015 for $16.7 billion.

No Time for Buyer’s Remorse

I’ve long been critical of Intel’s acquisition of Altera, as well as how poorly the management team at Intel managed the acquisition. Now we can clearly see how much value has been lost.

Silver Lake’s controlling interest in this division of Intel values the Altera-related assets at just $8.75 billion, which means that Intel will receive about $4.46 billion in cash for the transaction. By any measure, the acquisition of Altera has been a disaster.

Over the 10 years since the 2015 acquisition, about half the value has been destroyed.

But this is no time for buyer’s remorse. It would be stupid to triple down on Altera and try and force a way to make it work. This was a smart move by Lip-Bu Tan.

In 2024, the Altera assets generated $1.54 billion in revenue for Intel which is just 2.9% of total revenues. Not much at all. But worse, this division was responsible for a GAAP operating loss of $615 million last year. After a decade, Intel still couldn’t figure out how to make that business work.

Intel will deconsolidate the Altera-related financial results from Intel’s results later this year which will have a material contribution towards getting Intel back on track.

And aside from the improved financial performance, the even larger benefit will be to drive the much-needed focus in Intel’s business, the kind of focus that has been completely lacking over the last decade.

Private Equity

Some are questioning why Lip-Bu Tan didn’t sell the Altera division outright to another major semiconductor player. Here’s why Tan’s decision to go the private equity route was smart:

• First and foremost, speed. In less than four weeks since taking over, Intel and Silver Lake were able to agree on a valuation for the Altera division. That got the deal done.
• Silver Lake has a controlling interest in the business. That way, Intel doesn’t need to be responsible for the business at all, hence the ability to focus on Intel’s core businesses.
• Intel retains 49% of the division, preserving potential upside in a future sale of the business. Silver Lake, after all, is a private equity company, not a semiconductor company. Its “game” will be to increase the value of the Altera business over the next two or three years and then sell the company to another semiconductor company at a materially higher valuation. While Intel will never get its money back from the original acquisition, it is extremely likely that it will receive an additional $6–8 billion within the next three years when the business is sold.

Why am I so confident? Silver Lake isn’t new to the semiconductor industry:

• Silver Lake was part of a private equity consortium that acquired Agilent Technologies – which was originally Hewlett Packard’s semiconductor-related business – which eventually became Avago Technologies. And then Avago was merged with Broadcom which retained Avago’s ticker symbol AVGO.
• Silver Lake, through its relationship with Mubadala (the sovereign wealth fund of the UAE), invested in Global Foundries which was originally the semiconductor manufacturing arm of Advanced Microdevices (AMD). AMD spun out its manufacturing in 2009 in partnership with Advanced Technology Investment Company (ATIC) – a UAE-based investment fund that became Mubadala.
• Silver Lake was the third largest private equity investor in the consortium that took Philips Semiconductor private in 2006 in a leveraged buyout (LBO). That became NXP Semiconductors, my former employer. I was one of the senior executives responsible for turning that business around and ultimately taking NXP Semiconductors public in 2010. NXP Semiconductors turned out to be one of the most successful LBOs in semiconductor history.

Of course, considering that Lip-Bu Tan was a former board member of Intel, I expect that discussions with Silver Lake had occurred in the past as Intel considered all of its options to right the ship and get the business back on track. That would explain how a deal could come together so quickly.

And that’s not all…

TSMC to the Rescue?

Earlier this month news broke that it might enter into a joint venture with none other than Taiwan Semiconductor Manufacturing (TSM) to operate Intel’s semiconductor manufacturing business.

The rumor was that TSMC would take 20% of Intel’s manufacturing business, and that’s its equity stake would be earn with “in kind” investments.

That suggests that Intel wouldn’t raise cash from the transaction but would get “payment” in the form of sharing TSMC’s trade secrets and helping Intel with training so that it can return to bleeding-edge semiconductor manufacturing.

It’s not clear whether or not a deal will go through, but it is clear that Tan is considering all options to get Intel turned around.

One Down, Three to Go

This is refreshing to see, and it is the first time in a very long time that I can see some hope for Intel as a company.

As I wrote in The Bleeding Edge – Can Intel Be Saved? on April 2…

I’d recommend selling off the following:

• Intel’s Programmable Solutions Group – this is the field programmable gate arrays (FPGA) business that came to Intel through its acquisition of Altera in 2015. The acquisition was poorly managed and would do better at another large semiconductor company.
• Intel’s Accelerate Computing Systems and Graphics Group – this is the division that houses Intel’s GPUs. Intel missed the boat completely on GPUs and it is far too behind NVIDIA and AMD to catch up.
• Mobileye – Intel still retains ownership of 88% of Mobileye. Intel acquired Mobileye for $15.3 billion in 2017, and the company is now trading at a valuation of $10.3 billion. It is not core to Intel’s business – it should sell its stake in the company completely and raise capital.
• Intel’s Internet of Things Group should be spun out and sold. It is not strategic and not core to its business.

One down, three more to go. That’s a great start in less than a month.

That’s the benefit of bringing in an outside executive who is deeply familiar with the business. There’s no emotion for Tan around what needs to be done. No buyer’s remorse. No stubborn need to try and prove that an old acquisition might be able to work after all.

Just making the tough decisions, focusing on the business, and bringing it back to health as quickly as possible.

Lip-Bu Tan is on the right track.

Jeff