Hedge fund Third Point urges Intel to explore deal options

Option 1: Strip mine the company. This is more of the traditional PE/activist play. Once you stop spending all those R&D dollars, then Intel could extract a lot of profits from the x86 over the next decade or so. But this is only a short-medium term play. All mines run out of gold eventually.

Option 2: Split the Fab and design (x86) business. I'm not sure this helps that much, I think x86 is F'ed even if it could use the TSMC fabs. This is what AMD is already doing. The major important players (Apple, Amazon, MS, Goog) want to go ARM and then maybe RISC-V for their own strategic reasons anyway.

Option 3: Spend unheard of amounts of money to try and win back a process advantage over TSMC. This would entail spending tens of billions of dollars to basically make a bet that might pay off. It would be fun to watch, but success is not guaranteed, or even likely.

Is there an option 4?

Intel have now burned an enormous amount of capital through failed acquisitions (Mccafee, Altera, Nervana etc) - mainly because their attitude was "Oh we can do that too, and that's a really big market". They can't make FPGAs better than an FPGA company, they can't make security software better than a software company, they can't make AI chips better than a company that does AI.

What would make them successful again is to focus on their core business (a business they departed because the growth wasn't there). So that means spinning out fab (which let's be fair, will probably die), spinning out Networking, I was going to say spinning out FPGA but I just went and looked up the head of that group and he works at AMD now (unlucky for AMD). Spin out the data centre group. Spin out the IOTG group - who can then operate on reasonable scales rather than refusing to work on anything smaller than 100 million units. Spin out MobilEye (they fucked up the BMW project that got them acquired in the first place). Then what you're left with in Intel proper is just Martin from HR, and he was useless anyway.

There are any number of scenarios and playbooks certainly gamed out by investment bankers over the years that Intel could pursue.

What is lacking is a deep competitive drive driven by existential paranoia.

I am interested in seeing what Intel’s designers could achieve on TSMC 5nm, but that’s hardly the idea behind the deal here to secure on-shore chip manufacturing

TSMC has proven that the fab model works well, and Apple has proven a design model works well.

Intel could sell their design group to MS and open fabs to all. The u.s govt can require second source (eg spacex and ULA, envelope though the latter is more expensive) to help fund intels EUV transition.

This is the key point right here, and one our Government should be worried about. It's a national security issue if we aren't able to make our own chips.

It's a tough choice, because in the name of free trade you want to allow companies to build wherever makes the most sense cost-wise, but in the name of national security, you want to retain that critical skill here.

It's a tough balancing act.

As the RISC-V thing was getting going a few years ago I thought the commoditization of CPUs would make GPUs the next interesting IP. It seems the next thing is AI acceleration with Apple and nVidia ahead AFAICT.

Meanwhile RISC-V V is nearing completion and decent chips are starting to become available.

I dont see anything but fierce competition for Intel for the next 5 years or more.

1. They need to fend off AMD in x86 and in the absense of a process solution that means TSMC.

2. Opening up fabs would maximise the value of manufacturing and support ongoing investment as x86 declines.

3. Sell everything that is non-core and a distraction.

4. Fix the culture.

Note that 1. and 2. don't necessarily mean splitting the company.

Not simple though - especially 4. (easy from the armchair though!)

I would love to see Intel open up their fabs. More fabs in more locations would be beneficial to the chip world overall. However, I don't know if Intel can do it successfully.

The vertically integrated strategy pays off when you have enough volume in the money maker (CPUs) to justify the heavy investment in the bottom of the stack (fabs). Intel is nearly at a threshold where they risk losing hold of the CPU market.

I see about 4 paths. I don't know which one we're closest to.

1. The ideal future for everyone (except for their their competitors in the manufacturing space) is that Intel successfully spins out their fabs and we have a viable near-leading process available for the chip industry in the large. There's still a lot of money in not-quite-bleeding-edge processes.

2. There's a future in which Intel doesn't, they bounce back, and everyone calling for them to spin out their fabs looks like an idiot.

3. There's also a future in which Intel doesn't, they don't bounce back, and their process falls into irrelevance as they lose the volume to fund their fab development.

4. Then there's a future in which Intel does, they do or don't bounce back, and Intel's fabs are still not competitive enough to be used by anyone but Intel. It's still an open question whether Intel can reach the volume they need to make this pay off, for example. It's also not obvious that Intel's process, on a technical level, is ideal for anyone but Intel.

The ideal time for this to have happened was years ago when Intel still had a process advantage and could leverage that to bring in new customers despite the other risks, but it was precisely because they had a process advantage that they weren't going to do it. In fact, they went the other direction and let go of much of their custom foundry expertise, which was quickly scooped up by (now) more successful companies, and purchased firms that moved to the Intel process rather than maintain a customer relationship. Now they have to contend with the fact that they don't have an advantage and they aren't as accustomed to working with partners for manufacturing. It's an uphill battle, but it's doable. The geopolitics might be compelling enough that they can overcome this with pressure and funding from three-letter agencies.

Culturally, this might be better regardless of any of the technical details. I've often heard that manufacturing runs the ship at Intel and that they suffer from severe internal politics between orgs. Splitting up might be the right tool to reign all of those problems in. But, it might also be that Intel truly has been infected with the managerial MBA rot that technical folks like me like to complain about, or Intel has irreversibly shed its manufacturing talent. In that case, I don't know if manufacturing can be saved.

A random Intel device probably runs the software you want.

A random ARM device almost certainly doesn't run the software you want

Intel is failing to play this advantage.