Intel – Lifeline

Intel secures an anchor tenant.

• Intel’s arrangement with Amazon and the delay of its European rollout add to previous measures that aim to conserve cash underlying just how badly the turnaround has gone and the opportunities that Intel has missed.
• Intel provided two updates on 16^th September the first of which was an expansion of its relationship with AWS and the second a press release following a board meeting that aims to put an end to the speculation with regards to Intel’s future.
• Intel and Amazon are expanding their relationship such that Amazon and Intel will co-invest in custom chip designs for Amazon that will comprise of next-generation Trainium supported by a custom Xeon 6 chip.
• Trainium will be made on Intel 18A while the Xeon chip will be made on Intel 3.
• This is an important deal for Intel as it badly needs volume to fill its foundries, and Amazon will provide both high volume and also serve as a reference and testament that Intel’s foundry services are of good quality.
• However, it will not serve to dispel concerns with regard to the independence of Intel’s foundry where potential customers such as Nvidia, Broadcom, Qualcomm, MediaTek and so on will be concerned about being treated fairly given the in-house client.
• This is a concern, but it is one that can be dealt with as Samsung Foundry has been able to do for many years and so I am not convinced that Intel Foundry must be spun off to ensure its independence.
• This is why Intel is also announcing that Intel Foundry will become an independent subsidiary inside of Intel which will have its own independent board but continue to report to Pat Gelsinger.
• Intel is also being forced to focus on capital efficiency which is what the current 15,000 layoffs and the 2-year delay in European fab rollout are all about.
• Arizona, Oregon, New Mexico and Ohio continue as planned which keeps Intel in a position to receive further US government funding of up to $3bn.
• The sad part of this story is that much of this was avoidable.
• Intel has now missed two of the biggest semiconductor trends of the last twenty years (mobile and AI) and failed to sell Mobileye when it had the chance.
• Mobileye was purchased for $15bn and following its IPO rose to $33.8bn where it sat for a year before collapsing to its current $9.2bn.
• The writing has been on the wall with regard to Mobileye for some time as Qualcomm has steadily been winning and announcing contracts for the 2025 model year and beyond.
• Intel had the opportunity during 2023 to sell a good portion, or perhaps all of its stake in Mobileye which would have raised $20bn – $30bn of badly needed capital which could have been invested in the turnaround.
• This would have reduced the pressure on Intel and allowed it to pursue its turnaround strategy with a much freer hand.
• Instead, it has been forced into making cuts at a time when it should be investing to keep its competitors at bay.
• Intel is facing assaults on all fronts from Qualcomm and AMD in PCs and Nvidia and many others in the data centre and keeping share is going to be increasingly difficult.
• I have always said that at 10x PER, Intel is worth having a look at but the problem is that earnings estimates have fallen by more than the share price.
• The consensus for 2025 is $1.13 putting the company on 17.7x 2025 PER and there are plenty of others who are not in existential crisis, who are growing earnings and are trading at a lower multiple.
• Intel is not out of the woods yet and until I can see the way out, I am staying away.