News Intel Engaging in 'Semi-Destructive' Actions Against AMD, Says Firm

[GustavoVanni]

Then how do you explain this?

https://www.tomshardware.com/news/tsmc-might-cut-3nm-prices-to-lure-amd-nvidia

Sounds like TSMC is expecting to have more supply than demand for it, at the moment.

I think the articles on my post were written before everyone in this industry realized that in the next years everything will go down hill, with expected sales much lower than anticipated during COVID season 2020/21.

Then, again, I'm not an especialist, so I might be wrong about this whole thing.

 

[waltc3]

The problem for these people is that AMD is clobbering Intel in the server markets. For those who do not know, the server markets are where the real money is. The server market share is what counts the most, by a long shot. There's little Intel can do about that. I'm not worried about AMD, and "analysts" are frequently wrong, btw. In the CPU business, the server markets are the Holy Grail. Intel can ship a glut of products into the general PC markets, and it's just going to come back to hurt Intel! Guaranteed. Intel is fairly desperate, atm, if you haven't observed.

 

[Darkoverlordofdata]

Isn't setting strategic price points the whole basis of the free enterprise marketplace? I guess when the news is this ho-hum you need to make it sound exciting...

 

[Endymio]

Sure, you can find quibbles

A quibble? Your claim was that price-fixing cases required collusion to raise prices, when neither collusion nor raised prices are necessary. How could that statement possibly have been more wrong?

There's been a great wealth of economic scholarship, over the past century or so, and most of it is more worthy of citation than anything Rand ever said or wrote.

Rand wasn't an economist; why would you expect her to publish scholarship on the subject? Your statement is akin to claiming we should discount Steven Hawking's views on astrophysics, as he never published a treatise on brain surgery.

(And to head off the expected reply: no, Rand was not making a point of economics. She was exposing the tautological nature of the definition: no matter what course a company charts, they can be accused of unfairly competing. The economist would ask: "does such a definition increase or decrease the utility of resource allocation?")

Greenspan had a mixed history.

He's hardly unique in agreeing on this topic. Milton Friedman, for instance, also wrote extensively on the perils of overly defining unfair competition. Given he won the Nobel in economics, will you also denigrate him as being "unworthy of citation"?

There are few people I wont argue with on most topics discussed here, however, @Endymio is one of them, and econ is his schtick. He knows his subject matter and loads for bear. Usually disagreement with him is a disagreement with reality, at least on the topic of economics.

I appreciate the compliment. However, never fear: there are countless other topics upon which I am a thoroughly bloviating ignoramus.

 

[helper800]

I appreciate the compliment. However, never fear: there are countless other topics upon which I am a thoroughly bloviating ignoramus.

As you say, nobody is infallible, especially when it comes to multiple subject matters. Ill take your word in relation to economics though, pretty safe bet. I have already been humbled in that regard.

 

[Endymio]

Labor isn't the only thing in short supply. When any essential resource or service is in short supply (including things like fuel or shipping), then prices go up. When prices rise in a few sectors at once, it spreads throughout the economy and you get inflation.

Not quite. When a single commodity is in short supply, its price rises, but that very rise causes a reduction in demand for other goods and services (the so-called "income effect" in economics) which tends to reduce their prices in response. But when everything is in short supply, the net effect on prices is zero. (a variant of Walras's Law).

That's in a long-run analysis. In a short-run analysis, consumers can reduce their PSR (savings rate) and partially negate the income effect. That can produce short-term inflation, but its cumulative effect on inflation is zero (inflationary pressure today is countered by deflationary pressure tomorrow).

As you've so thoroughly castigated Rand, I'll quote Friedman: "Inflation is, always and everywhere a phenomenon of monetary policy." The law of supply and demand applies to currency no differently than anything else. When you double the supply of dollars, the only thing that can prevent inflation is doubling the amount of goods and services upon which you can spend those dollars.

 

[helper800]

Not quite. When a single commodity is in short supply, its price rises, but that very rise causes a reduction in demand for other goods and services (the so-called "income effect" in economics) which tends to reduce their prices in response. But when everything is in short supply, the net effect on prices is zero. (a variant of Walras's Law).

That's in a long-run analysis. In a short-run analysis, consumers can reduce their PSR (savings rate) and partially negate the income effect. That can produce short-term inflation, but its cumulative effect on inflation is zero (inflationary pressure today is countered by deflationary pressure tomorrow).

As you've so thoroughly castigated Rand, I'll quote Friedman: "Inflation is, always and everywhere a phenomenon of monetary policy." The law of supply and demand applies to currency no differently than anything else. When you double the supply of dollars, the only thing that can prevent inflation is doubling the amount of goods and services upon which you can spend those dollars.

I am curious, how do you perceive the CHIPS Act and its overall impact on the industry?

 

[TJ Hooker]

Total population of working age vs the percentage of that population that actually works minus those with a physical disability or in school.

Current percentages only capture those were working within the past year, anyone who hasn't had a job for a year or more completely drops off the statistic.

They only drop off the stats when haven't even looked for work for a year. Not merely that they haven't had a job for a year. The "marginally attached to the labor force" demographic that BLS reports.

But yeah, I guess it would be nice to see, of all the people who could be working, and aren't otherwise occupied, how many are?

There is employment population ratio to look at. If we assume that the proportion of people who are disabled, in education, stay at home parents, retired, etc. haven't changed too significantly, we can use it as a rough proxy to at least see changes in 'actual unemployment'. That's a pretty large assumption though, I admit.

 

[palladin9479]

They only drop off the stats when haven't even looked for work for a year. Not merely that they haven't had a job for a year. The "marginally attached to the labor force" demographic that BLS reports.

But yeah, I guess it would be nice to see, of all the people who could be working, and aren't otherwise occupied, how many are?

There is employment population ratio to look at. If we assume that the proportion of people who are disabled, in education, stay at home parents, retired, etc. haven't changed too significantly, we can use it as a rough proxy to at least see changes in 'actual unemployment'. That's a pretty large assumption though, I admit.

The way U6 report works is that it assumes everyone that isn't employed is "looking" for a job and that they either have found one or are no longer "looking" after 12 months. They literally get dropped off entirely are don't count on either side of the line. Yet those people still exist, they still need food, clothing, shelter, they still require societal resources.

The only people excluded should be those physically disabled, in school or too young / old (under 18, over 70). Everyone else in the population needs to accounted for. The point is to get a good idea of the number of healthy working age individuals that are and are not employed so that policy and law can be set to better the economy. Playing around with numbers to make the result look good is how we got fake inflation rates in the first place.

 

[TJ Hooker]

The way U6 report works is that it assumes everyone that isn't employed is "looking" for a job and that they either have found one or are no longer "looking" after 12 months. They literally get dropped off entirely are don't count on either side of the line. Yet those people still exist, they still need food, clothing, shelter, they still require societal resources.

The inclusion criteria for "marginally attached to labor force" (and by extension, U-6) mentions nothing relating to time-since-last-job. It's merely everyone who: is 16+, not currently employed, and has looked for work in the last 12 months but not within the last 4 weeks.
https://www.bls.gov/cps/definitions.htm#marginallyattached
"In response to survey questions, people marginally attached to the labor force indicate that they have searched for work during the prior 12 months (or since their last job if it ended within the last 12 months), but not in the most recent 4 weeks."

Unless I'm misunderstanding you, your statements do not match BLS definitions. So the only people effectively left out of the stats are those who are capable of working (and of working age) but have not even looked for a job in over a year.

Edit: Small clarification, as pointed out below, people have to be "actively" looking for work within the last year to be included. E.g. skimming public job boards without ever sending out applications wouldn't count.

 

[palladin9479]

The inclusion criteria for "marginally attached to labor force" (and by extension, U-6) mentions nothing relating to time-since-last-job. It's merely everyone who: is 16+, not currently employed, and has looked for work in the last 12 months but not within the last 4 weeks.
https://www.bls.gov/cps/definitions.htm#marginallyattached
"In response to survey questions, people marginally attached to the labor force indicate that they have searched for work during the prior 12 months (or since their last job if it ended within the last 12 months), but not in the most recent 4 weeks."

Unless I'm misunderstanding you, your statements do not match BLS definitions.

Because the questions of the survey aren't worded that way. To be actively seeking you have to be submitting resumes, attending job fairs and so forth. People tend to do that for a few months after losing a job, then they kinda stop. They are still able and available to work if pressured but the "new norm" is not working. They end up being categorized as "not participating in labor force".

 

[helper800]

The inclusion criteria for "marginally attached to labor force" (and by extension, U-6) mentions nothing relating to time-since-last-job. It's merely everyone who: is 16+, not currently employed, and has looked for work in the last 12 months but not within the last 4 weeks.
https://www.bls.gov/cps/definitions.htm#marginallyattached
"In response to survey questions, people marginally attached to the labor force indicate that they have searched for work during the prior 12 months (or since their last job if it ended within the last 12 months), but not in the most recent 4 weeks."

Unless I'm misunderstanding you, your statements do not match BLS definitions. So the only people effectively left out of the stats are those who are capable of working (and of working age) but have not even looked for a job in over a year.

These statistics do not include those who work for cash, and all the implications that has.

 

[bit_user]

Simple math hasn't been provided, as quite a few people have pointed out already. There's no evidence of the actions, or premise that it's hurting earnings. As such legitimacy has not been established.

I'm not disputing that. I'm just arguing that there's indeed a basis on which they can judge that it's being destructive.

As for what evidence they've provided, such research firms have access to non-public information. They can't and won't show all of their work, when they issue these kinds of statements. They basically trade on their reputation. You're free to ignore their advice, but perhaps it's at least worth considering.

 

[TJ Hooker]

They are still able and available to work if pressured but the "new norm" is not working.

Ah, I should have figured you had something like this in mind when arguing that the official job numbers aren't the 'real' ones.

What evidence do you have that there is a significant, recent increase in the proportion of people who are able and available to work, but simply choose not to?

 

[Endymio]

What evidence do you have that there is a significant, recent increase in the proportion of people who are able and available to work, but simply choose not to?

The labor force participation rate has dropped from 63.3% pre-pandemic to 62.3% -- a full percentage point difference. Given unemployment rates are equal between those periods (3.5%), there is no other explanation -- unless one postulates 1% of the labor force losing their able-bodied status.

I am curious, how do you perceive the CHIPS Act and its overall impact on the industry?

I'll give three takes, cover all bases, and thereby ensure I can't be proven wrong. From a strictly economic perspective, government interventions of this sort always decrease total utility. Subsidizing an industry doesn't "lower costs and create jobs", as some of the bill's sponsors claim.

However, such pure analyses don't account for wars, pandemics, nor trade embargoes for political goals. The subsidization of certain critical industries -- agriculture, energy, and, in our modern world semiconductors -- could be seen as a worthwhile insurance policy: wasted money spent on the policy premiums ... until you cash in a disaster claim. Given Taiwan's geopolitical situation, that's not wildly unreasonable.

The third take is that, on a pragmatic basis and given the size of the industry and costs thereof, I suspect the size of the CHIPS Act to be substantially too small to make a decisive impact, perhaps not even a measurable one.

 

[bit_user]

She was exposing the tautological nature of the definition: no matter what course a company charts, they can be accused of unfairly competing.

I don't think it's so enlightened or enlightening. It's just pandering to self-pitying businessmen who lament that they're not rich and loved by the masses.

Whenever interests aren't perfectly aligned, neither will semantics.

More to the point, business is brutal, and there'll always be someone whinging if you're doing it well. Such complaints don't warrant time or attention. What is notable is when laws and regulations are breached, or investors' interests are compromised.

Milton Friedman, for instance, also wrote extensively on the perils of overly defining unfair competition. Given he won the Nobel in economics, will you also denigrate him as being "unworthy of citation"?

Of course he's worth studying, though he hardly has the final word in all things economics.

 

[bit_user]

When a single commodity is in short supply, its price rises, but that very rise causes a reduction in demand for other goods and services (the so-called "income effect" in economics) which tends to reduce their prices in response.

Demand for energy is fairly inelastic, meaning that demand-reduction elsewhere can't easily cancel it out. Same with some of the other critical resources I mentioned, like helium, fertilizer, and grain.

One idea behind raising interest rates is demand-destruction, since reducing the amount of borrowing can lessen demand across a broader swath of the economy.

As you've so thoroughly castigated Rand, I'll quote Friedman: "Inflation is, always and everywhere a phenomenon of monetary policy."

That's too absolute. You can see exceptions to this in countries with tight monetary policy (as typically mandated by recipients of IMF loans) that are experiencing inflation due to rising prices of commodities they import.

 

[bit_user]

I'll give three takes, cover all bases, and thereby ensure I can't be proven wrong.

​

"Give me a one-handed Economist."​

--Harry S. Truman​

​

Subsidizing an industry doesn't "lower costs and create jobs", as some of the bill's sponsors claim.

Depends on what strings are attached. Also, the total number of jobs might not be substantially impacted, but it's where those jobs are that counts.

The third take is that, on a pragmatic basis and given the size of the industry and costs thereof, I suspect the size of the CHIPS Act to be substantially too small to make a decisive impact, perhaps not even a measurable one.

The way I look at it is that it helps ensure domestic semiconductor production capacity is nonzero. It's a lot easier to scale up an industry from something, than to build it from nothing at all. So, even though it won't keep us from being dependent on Taiwan, it'll lessen the amount and duration of pain, if/when that supply gets cut off.

 

[zavalita]

"Breaking news! Company reduces prices to be more competitive!" First people complain that Intel's prices are too high. Now they're complaining they're too low? If Intel becomes unprofitable, perhaps then there's something to talk about beyond just the short-sighted impact on stock price.

The stock brokers didn't "complain", readers of Tom's Hardware didn't complain. Reading comprehension ?!

 

[KyaraM]

The only people excluded should be those physically disabled, in school or too young / old (under 18, over 70). Everyone else in the population needs to accounted for. The point is to get a good idea of the number of healthy working age individuals that are and are not employed so that policy and law can be set to better the economy.

Mental diasbility and mental illness can also prevent you from working, and can also lead to you to be classified as diasbled and elligible for benefits despite being physically completely healthy. Meanwhile, physically disabled people can still, in many cases, work (think someone in a wheelchair doing a desk job), and do so. You are generalizing a bit too much here.

 

[jasonf2]

AMD did build their own fabs, then AMD spun off their fabs to make Global Foundries in 2009, kinda like what Intel plans to do with their fabs. Only difference is I doubt Intel will completely divest over time their new foundry entity like AMD did in 2012. Kinda makes you think though, if AMD retained their equity in Global Foundries, they probably could have vetoed the decision to abandon advanced process nodes smaller than 12nm and would still be able to utilize Global Foundries services.

There is a pretty dramatic difference between what AMD did with GlobalFoundries and what Intel is doing. AMD was not in a great place when all of this went down. When they split out global foundries it not only created a cash infusion for AMD but also let GlobalFoundries expand out as a freestanding fab operation. GlobalFoundries is still very much alive and kicking, but they decided not to pursue bleeding edge fabs, instead focusing on the more mature tech used in a huge amount of daily use stuff. Think coffee maker controllers and auto ECMs, not gaming computers. If you look at the exponential cash flow requirement increases necessary to maintain the latest and greatest nodes AMD saw the writing on the wall clear back in 2008 and I cannot blame GlobalFoundries for going after the mature node market.

Intel on the other hand today is looking to become TSMC. Their foundry as a service model is all about them selling their foundry services to others not about divesting themselves from the business of making chips. Again with the crazy R&D expenses necessary to keep up, creating other cash flow streams are just good business. In today's fabless world if Intel isn't playing in this space they will find themselves locked out of new emerging markets. The smartphone game and TSMCs rise to dominance is a great example. Qualcomm cannot simply rebrand an Intel chip and sell it. But Qualcomm could have their chips made in Intel fabs. So if Intel only makes chips it designs they have to not only head to head designs and market them, but it also has to bear all of the Fab R&D to do so at their margin and market share. When you have the dominant market share position Intel has enjoyed in the past this isn't a problem. But TSMC isn't just playing in the CPU/Server market. They are making GPUs, CPUs, Cellphone SOCs, and anything else that someone will pony up the cash for. Many (if most) of these markets are mature enough that there simply isn't enough market space for Intel to muscle proprietary chips in despite efforts to do so. The next best thing for Intel, and the most probable, is to maintain top tier fab nodes and sell those services to companies like Nvidia, Apple, Qualcomm and (as longshot as it may seem) even AMD. This keeps them in a place where they can afford to stay on top of the Fab Tech, and also have access to their own manufacturing.

Again as I said earlier what AMD did and where Intel is today is kind of a apples to oranges comparison. Also AMD at it's high point had something like a 34% stake in GlobalFoundries and was itself heavily leveraged by ATIC. That dwindled to around 14% at the end and the whole divestment thing was pre agreed on happened in ~2012. GlobalFoundries didn't make the focus changes until ~2018 and at about the same time AMD was able to amend it's wafer supply agreement with them to allow access to other vendors because GF dumped bleeding edge. So in hindsight GlobalFoundries node path choice led to AMD being able to work with TSMC which allowed them to have the rebirth they have enjoyed for the last couple of years. As a point of personal opinion if they had stayed with GlobalFoundries there is a pretty good chance that AMD would be gone right now. Without TSMC production nodes Ryzen would have been impossible and all GlobalFoundries was doing was licensing Samsung tech and holding AMD back in a market that Intel owned.

 

[bit_user]

Intel on the other hand today is looking to become TSMC. Their foundry as a service model is all about them selling their foundry services to others not about divesting themselves from the business of making chips. Again with the crazy R&D expenses necessary to keep up, creating other cash flow streams are just good business.

Yes and no. IFS is trying to become TSMC. Intel will spin them off, once they become self-sufficient.

Why? Because the fabs are a big, capital-intensive boat anchor dragging down Intel's profit margins, and Wall St. doesn't like that sort of thing. They would prefer to see the IP part of Intel get untethered from the financials of maintaining and running fabs, and also freely able to use whichever fabs happen to be best/cheapest at the time. It's also better for IFS, as there aren't the trust issues for someone like AMD or Nvidia using it under independent ownership that the would be while it's still part of Intel.

if they had stayed with GlobalFoundries there is a pretty good chance that AMD would be gone right now.

Yeah that, or some other bad outcome. What they did probably worked out best for both companies.

 

[jasonf2]

Yes and no. IFS is trying to become TSMC. Intel will spin them off, once they become a separate entity.

Why? Because the fabs are a big, capital-intensive boat anchor dragging down Intel's profit margins, and Wall St. doesn't like that sort of thing. They would prefer to see the IP part of Intel get untethered from the financials of maintaining and running fabs, and also freely able to use whichever fabs happen to be best/cheapest at the time. It's also better for IFS, as there aren't the trust issues for someone like AMD or Nvidia using it under independent ownership that the would be while it's still part of Intel.


Yeah that, or some other bad outcome. What they did probably worked out best for both companies.

If Intel spins them out there is still a significant difference here. AMD was tanking, couldn't find a real buyer, and what they did was sell their fab operation off to ATIC through a shell move. Intel may spin IFS out as an entity, but will still maintain control of the entity and make money doing it. These type of liability shells are constructed all of the time in business where the holdings company still retains ultimate control. AMD's deal was really a highly leveraged shell game selloff of the fabs to ATIC that included the transfer of over a billion dollars of AMD debt to the new company along with direct capital infusion in AMD in the form of stock purchase and even then they had to remain minority partners with a pre agreed exit and locked vendor agreement. I am not going to say that Intel isn't going to sell stock out in the IFS shell, but I would be seriously amazed if they let more than 49.9% of it go and it won't be a leveraged deal. I think IPO rather than the fire sale that AMDs fabs were by that point.

 

[bluvg]

The stock brokers didn't "complain", readers of Tom's Hardware didn't complain. Reading comprehension ?!

Where have you been? Back before AMD took the performance crown most recently, it was a very, very common complaint that Intel offered processors that, while the fastest, were overpriced.

 

[bluvg]

First, you're confusing two different points of view. Consumers want low prices, while investors want high margins (which go along with higher pricing).

I wasn't out to write a novel, but perhaps I should have clarified those complaints have been on the consumer side. I take it as fairly obvious that consumers and investors are looking for different things. It's also pretty clear that investment value stock-price-wise is hardly in lockstep with company performance.