For its big comeback, Intel needs to spend money, and it's making less and less

Intel may not ever fail, but that sure doesn't mean shareholders will reap any rewards. Boeing won't fail either, but has been losing huge chunks of money every quarter for 2 years now.

Won't fail but their stock can and likely will continue to fall for quite some time.

When its self-professed peer state adversaries are Russia and China, is it surprising the US is bending more towards national champions?

The CHIPS and Science Act is nowhere near a qualitative shift away from the investor-centric mentality of managing a public company. Many investors are short-term greedy and don't learn much about the real goings-on of the company. It makes a few things cheaper, but if you compare to TSMC and Samsung it's quite clear that Intel doesn't get anywhere near the amount of government money or political power that its competitors at the leading edge happen to have in their respective localities.

That would destroy the stock price, and executives get paid largely in stock.

ASML is a bigger single point of failure, since they make the machine most critical to any modern fab from Intel, TSMC, or Samsung.

You mean like ASML already is? They are the only manufacturer of EUV lithography equipment for fabs.

https://analyticsindiamag.com/is-asml-the-most-important-tec...

I model it like a sovereign default risk, only the sovereign is a corporation.

There is little evidence inflation below a huge 40% or whatever causes any serious problems.

Hard to say.

I am seeing the service economy wracked by a failure to invest in people, process and technology. Go to a restaurant, hair salon, pharmacy, or attempt to ride the bus (the one I ride in the morning has been canceled) and it is right in front of you.

I know just enough process science to be dangerous, but something I see in all of these cases is very long queues which themselves become a source of disruption and chaos. People seem to think that being willing to wait in line is a sign of maturity, but it is reduced to practice people seem to almost universally not understand that your doctor can't see more patients by making them wait longer.

I was at a Dunkin Donuts the other day and rather shocked that, because of a lack of hash browns, the kitchen was unable to serve any customers waiting in the drive through for almost 30 minutes.

The anti-inflationist would say the problem is that people have money in their pockets to order hash browns that the restaurant can't serve and if we crashed the economy hard enough demand would collapse below supply. My fear is that, without the capital being there to invest in people, process and technology, supply will collapse and we'll be stuck with an economy where "nothing works" but the level of demand keeps dropping.

There is no evidence that inflation under 40% poses serious economic problems.

There is no evidence that the current inflation is caught in something inertial like a wage-price spiral.

The reason we have a chip shortage in the first place is that slow growth has rationalized underinvestment. I suppose we can have a reshored chips industry by handing out electronics coupons in a recession, but that is a really blunt instrument with high distortion risk.

It's not like all chips are fungible or whatever, there is a real risk of subsidizing the "wrong thing" without a broader macroeconomic story (which could be planned in theory, but won't be in the US so we must rely on markets) to validate the "end use values" --- ensure the downstream benefits are extent to discipline the upstream chip production why also incentivize it.

The good thing about finding bottlenecks in a high pressure economy is those are the actual problems to be solved. But problems shift --- the goalposts actually aren't fixed --- so better to keep things high pressure the whole time.

Monetary policy acts on aggregate demand with long and variable lags. This is the consensus position. The latencies are comparable, with plenty of uncertainty about each.

If you are the Fed in one hand, and CHIPs in the other hand, it is like being in a pool of molasses trying to make a waves in a long jump rope that don't interfere once they meet in the middle.

> But inflation corrections can happen in a year or two.

Monetary policy sucks, there is little evidence of it ever creating a soft landing. (As opposed to regular market forces incentivizing more supply and fixing the issues on its own.)

Also, the consensus view is it takes up to 2 years just to come in effect! The Fed is probably creating a 2023-24 recession right now just as the Ukraine war and other sources of supply disruption stop being a problem!

>The truth is that all semiconductor vendors are heading for a future where they're selling an undifferentiated commodity product.

This is what I'm hoping for as well. But there are huge upfront costs (>20b). This will limit the number of potential players.

The differentiator in the server space comes in power management and thermals, all things that AMD does better than intel. AMD also has better yields and better margins because of their chiplet design philosophy.

ARM also has a significant thermal and power edge.

Intel is chasing the top, but actually they should be thinking about efficiency.

I mostly agree, and the compaction of value accrued at the lower layers of the stack are a real thing, but there are two wrinkles to the industry that I find interesting. First, like advanced turbine engine technology semi development continues to get more expensive as it matures and that drives consolidation rather than proliferation. Second, for now the end of Moore's law for companies with high pricing leverage looks like a move from "more for less" to "more for more" where they use performance and vertical integration to create a moat around their part of the market. You can see this in Apple shipping chips that are multiples larger than previous laptop processors, or NVIDIA pushing performance up but jacking prices up a the same time. Intel is going to have a hard time putting a 400mm2 chip into a Dell laptop at pricing and margins everyone can live with. At the same time they're going to have a hard time making margin on Xeons and still being competitive in TCO terms with whatever the Annapurna team can cook up for AWS using ARM IP on TSMC. This is all a roundabout way of saying unlike say the early PC industry I don't think maturing is going to level the field or lower prices, actually the opposite.

The problem is that Intel competes with Apple, Nvidia and AMD. They have a conflict of interest, just like Netflix has with content producers, and that's going to cause issues for their growth.

Hard to tell, TSMC's faster cadence means you'd have to get a snapshot of total output on N7 and newer but I don't think they break that out. They report numbers here and there in the press but usually only for the new process they're promoting and for example a lot of N7 from two years ago probably got cannibalized by N5 etc. For N5 and newer Intel doesn't have any equivalent process, and it sounds like they will be doing some products on TSMC at N3? If you did a histogram of wafers by effective feature size I think the shapes for Intel and TSMC would be very different but we know TSMC has at least two big customers selling N4 product today (Apple and NVIDIA).

> when paired with a 4090, the intel CPUs take the lead

Crypto mining is out, but enterprising gamers can now defray their costs by operating a steam turbine with waste heat.

> They win absolute performance on Desktop CPUs but lose on perf-per-watt.

How is that possible? They have better thermal design or something?

right, pair your $400 CPU with a $1400 video card, and man, look how fast it is for graphics applications!

I've seen this in action professionally and it drove me up the wall.

At a previous job I had to spec out some hardware procurement on the order of a million dollars over two years, and all of my compute was based around AMD EPYC. The amount of fighting and arguing that this resulted in was wild. These were all still supermicro boxes, still from the same trusted vendor we were using, but I had to spell out many times that we were basically getting double the compute for what we would have got otherwise.

Worst part was when the first shipment arrived and nothing booted and worked and turned it out be a bad batch of drives from micron that had bad controllers, the same ignoramus used it as a "I told you so" moment, even though the vendor confirmed it was a micron issue and immediately shipped over replacements.

Did you not try Windows on M1?

AMD doesn't even have a process do they? So it's presumably about x86_64 and associated IP.

Most servers don’t need raw performance from a single chip they can do their workloads by horizontally scaling.