I’d look at this a slightly different way, the supply issues are the main thing helping Intel retain market share - the longer the supply issues last the better it is for Intel.
Everyone else is battling for capacity at Samsung, TSMC - but because Intel has their own fab they can produce as much as they need. This is good because even though their products are generally much worse (especially on the DC side) customers are still buying them because the alternative is a 3 month wait from AMD.
Really the ideal situation for Intel (and what I’m holding out for) is:
The supply constraints extend as long as possible (2023+)
Diamond Rapids (or in an ideal world: Granite Rapids) represent a big jump over Genoa
Process node advancement keeps following the roadmap
I think a turnaround is certainly possible, it’s just still a bit expensive for me to take that gamble right now.
[TradingView - AMD posting a 17x return, outperforming nearly 38 times the return that Intel gave, excl. dividend pocket money]
AMD has been the bearer of great returns for many investors providing a 17 fold return on investment in just 5 years. A reasonable PE ratio of around 37, not dramatically outpacing the S&P500 benchmark everyone loves, that tells us that AMD still has growth ahead.
I could dive into fundamentals and share charts of future growth of the technology sector but I think Intel could be the better bet now… why?
Simply because a China-Taiwan crisis produces asymmetrical risk that cannot be ignored!
[Image from Foreign Affairs - Although war is not a certainty, nor wished for by any reasonable person. China has made its desires known. Even physical embargoes and a drawn out cold war could cause dire disruptions to market that China is currently hedging itself against, with domestic production growth in fabs.]
To be clear, AMD uses TSMC for all its production. It does not produce any chips itself.
Of the 17 TSMC fab locations, all except 1 are located within either China or Taiwan. That is the TSMC WaferTech Fab in the USA that produces 160 nm chips that are of no use to AMD.
Intel on the other hand has virtually all its manufacturing on US soil for the x86 architecture that competes directly with AMD. Furthermore, Intel is venturing into the GPU market and although this is being outsourced to TSMC currently, Intel has the capability (possibly not the capacity) to produce this in-house.
So when the Pentagon realises its mistakes, and China pushes for more domestic production of chips, we really have to ask… can we hold AMD through that or is Intel the better bet?
To wrap up this thought, please just pretend that RISC-V cannot slide into these markets to dethrone AMD and Intel as the kings of this market should the s**t hit the fan.
Sorry AMD, time to reduce my exposure.
A tiny up turn for tech stocks yesterday but not for Intel. I’m still 10% down on this. Full year earnings beat the year befores at $79 billion but net income was slightly down at $19.9 billion although their profit margin has been hit this has been more to do with higher costs of things
Tour/explainer video here about one of Intel’s fabs (Fab 28) in Israel (Linus Tech Tips video, unfortunately the title doesn’t make it clear when sharing the link…) - I Can Die Now. - Intel Fab Tour! - YouTube
Gives an idea why these fabs cost many billions to build and purchase all the machines.
Quite a few parts get blurred out with the secrecy around it, but the bit about polishing and it’s precision being down to atoms is mind boggling.
Intel sounding confident about progress on their 20A and 18A nodes
And with TSMC indicating production of 2nm in 2025, both these and Samsung might be at same 2nm process technology around 2025
IMO - See how it goes, as I’m not fully convinced because of Intel’s past struggles to deliver 10nm.
TSMC have the better track record delivering new process nodes.
If Intel can deliver 4nm roughly on schedule it’ll look more convincing.
From the earnings call, they seemed to be saying the operating margins were being impacted by the production ramp up of new CPUs and GPUs, they seemed more positive about the second half of the year with most of the new products getting out then
Intel is concentrating on long term growth as part of its strategy, it will be making a series of product and technology announcements at Intel VisiON 2022 on 10th and 11th May. Lets see what that brings?
Its not growing and people won’t price the stock at a premium until they are able to prove that the foundry business will be profitable and give the company some growth. Right now AMD and NVIDIA are taking up market share causing INTEL’s margins to also drop, so its been hard for investors to justify buying in despite the cheap valuation. I think the foundry business will be a strategic play that the US govt will support and so at under $45, I’m a buyer of INTC. I’m holding this until around 2028.
I think the the only real change is that they are further extending some of the ramp delays they mentioned in Q1.
I’m not getting any more just yet, the entire value of Intel hinges on this turnaround story materialising. I bought some (at about this price) based on their vision with the plan to continue to buy in as they successfully execute on their milestones. The problem is that even after big delays they are still stumbling with SR ramp being delayed yet again.
I don’t think this delay matters much in isolation (SR is an entirely uncompetitive product anyway) but they have so many big milestones to hit (across Architecture, Gate design, Process, Lithography, Packaging) by 2025 to reclaim leadership that they really can’t afford to falter on any of them.
