Yep. Slowly and all at once.

Especially if you ignore or avoid the evidence in front of your eyes.

Then why am I having trouble getting the GPU capacity I need?

GPU scarcity is a really poor proxy for the health of the AI market. I don’t know if I should be surprised that someone so bullish about it doesn’t understand this.

you should not be surprised. paper millions already borrowed against at stake

Did you read the OP? Because Nvidia gave money to people who aren't you, so those people could give the money back to Nvidia for GPUs. In other words they gave away all of their GPUs in exchange for stakes in startups, because (1) they could not sell that quantity of GPUs at market prices ordinarily and (2) they were betting that VC hype and the perception of high demand generated from doing so would be worth more than selling GPUs normally.

But presumably those gpus are being actually used by those companies? Even if their prices are artificially inflated? Or are these buyers sitting on piles of unused gpus to make demand look good?

I mean even Satya Nadella famously said that Microsoft has plenty of GPUs just nowhere to put them. There are definitely plenty of unused GPUs out there. Jensen is also a big fan of scarcity (said in multiple interviews) so obviously he’d want to keep it up at all costs.

> These GPUs are probably in use now,

A different question might be, will these companies grow over the new few years?

If I give you 10 Million GPU in exchange for 25% of the company and the company goes bust or has negative growth (and thus is forced to sell assets), effectively I just put 10M GPUs on the market with a huge discount.

At a grand enough scale, this undermines the GPU marker and one could argue that Nvidia has been doing this on a massive scale.

> Or are these buyers sitting on piles of unused gpus?

Yes, they are literally doing exactly that. Noting that they aren't truly "buyers" so much as "gift recipients" (in exchange for ownership stake, but the value of their remaining ownership also goes up if these gifts generate sufficient hype for investors to invest more). GPU production is outstripping datacenter capacity.

> But presumably those gpus are being actually used by those companies?

But, even if they were all being used, this wouldn't mean much. The central point is: suppose Nvidia has produced 10 million[1] datacenter-class GPUs during the LLM bubble. Nobody can afford to, or is willing to, buy 10 million datacenter-class GPUs because the economics don't work out. But if they gift 10 million GPUs, and frame it like they sold the GPUs (a framing that is effective -- although this has been an open secret, this is being reported like news by the WSJ, and to the majority of their readers, it is news), it makes Nvidia GPUs appear to be extremely desirable. Even though, logically, an item that you have to give away because it's not economically worth its market price is the exact opposite. The fact that you can use goods that you've effectiely been given for free does not increase the actual economic value of those goods.

[1] (hypothetical number, it's vaguely around this order of magnitude)

Why would they sit on piles of unused GPUs when they could rent them out or use them? Even if they were free that is still not what a business would do with them.

"Why don't they just use their unused GPUs"? - because they literally cannot. GPUs require electricity. They have more GPUs than they can supply electricity to.

https://www.datacenterdynamics.com/en/news/microsoft-has-ai-...

> Microsoft CEO Satya Nadella has said the company has AI GPUs sitting idle because it doesn’t have enough power to install them

https://www.businessinsider.com/amazon-nvidia-ai-gpu-data-ce...

> Some AWS sites had room for more racks but not enough available power to turn them on

> At least one AWS team did not have "a clear forecasted path" for getting the incoming racks into data centers at the time

Do you truly believe that the cost that can be charged for inference is below the cost of electricity?

No. I know for a fact that the take you've stated is incorrect, because I work for a startup that makes 8-digit profit annually on inference, with zero outside investment. I don't know why you've latched on to my comment to put this take into my mouth, because I never said anything resembling it. It can be true that a technology has some use while also being true that there is a bubble around people believing things about the technology that aren't true. The internet was one such example: the dotcom bust did not indicate that internet services were not financially viable, but that there were highly irrational allocations of investment that resulted in a lot of people losing a lot of money.

There is a fundamental difference from being able to profit on running some number of GPUs, and being able to profit from running 1 million GPUs. We do not have a business case for 1 million GPUs. Our startup's premise is that there are economically viable niches to use LLMs as tools, which is in stark contrast to OpenAI and Anthropic who are at least publicly selling the idea that LLMs are a form of genuine intelligence that will replace all human labor if you just give them another $10 billion dollars, and then another $10 billion dollars, and then another $10 billion dollars, .....

You may not have a business case for 1 million GPUs but you are not the only business. You clearly have a decent business case for some level of inference as do many others.

The gap between how many GPUs people have a profitable business case for and how many are actually out there is, however, a problem. Per their revealed financials for the pending IPO, Anthropic is deeply, deeply in the hole relative to their spending commitments, which are nearing $1 trillion over the next 5 years. OpenAI's do not appear to be much better from the more limited information we have about them.

The bubble popping will actually make business cases for GPUs much more common to come by, when businesses can acquire them for closer to their cost of production rather than at the insane markups Nvidia currently lists them for. But that won't do Nvidia much good, because Nvidia's valuation is predicated on being able to "sell" as many GPUs as they can produce at a 10x markup over the cost of production - even if much of what is being "sold" is really being given away for equity in businesses that are not financially viable.

> The gap between how many GPUs people have a profitable business case for and how many are actually out there is, however, a problem.

Is this a hypothetical gap or do you have actual numbers to back this up?

The rest of that paragraph continued on to mention two of the biggest users not actually having a profitable business case for the amount of GPUs they have.

That is a huge logical leap you are making then and it isn't supported by the evidence you gave. Are you suggesting that that two companies having more GPUs than they can use is proof GPUs aren't useful at their current price point?

There is a limit to how many GPUs can be profitable at this price point. Suppose you're running a SaaS. You might be able to charge $20/mo and get a sufficient number of customers to cover X amount of GPUs. This does not mean the market can support X * 100 GPUs profitably. THere is a limit to what the market can bear. THere are not infinite customers, in any niche, who will pay $20/mo for an LLM-based SaaS. At some point you will saturate your market of potential customers and purchasing more GPUs will not benefit you. And this even applies cross-SaaS. There is a finite pool of any amount of customers who will pay $20/mo for LLM services. If someone is subscribing to a competitor, they are less likely to pay $20/mo to you as well. All of this to say, it is certainly possible to oversupply GPUs that have a 10x markup. If the price is closer to production cost, then maybe they're cheap enough that you can offer a $10/mo plan profitably, and reach a much wider pool of customers and benefit from more GPUs. But there is a finite number of GPUs at which point the market for a sustainable business cases is satiated.

(Consumer sub prices were used for illustrative purposes but this obviously applies to token and enterprise spending as well. Although they're being a little loose due to FOMO, enterprises will not sign blank checks forever if the promised "replace all human labor" does not materialise)

Also, it goes without saying that a GPU that can not even be powered on for lack of power capacity is not being used profitably, at any price point. It could not be more apparent there is an oversupply relative to its current market price, which is a direct result of Nvidia giving away boatloads of GPUs for stakes in unprofitable companies.

OpenAI is in much worse shape because at least Anthropic has the growth to try to justify the spend. I can't see how the economics work but it is at least in some kind of margin of error.

OpenAI just seems like a black hole for money.

I think what you are describing is the sociological/ecological function of a bubble. We pull an unreasonable amount of demand forward, someone else pays for it and when the bubble implodes the good companies that survive the stress test get a free lunch.

Best of luck at the start up. That all sounds right to me. Real intelligence is specialized and ecological. The grand irony to me is that I would love to be an organism that can walk the way I do but also be able to fly like a bird and swim like a shark.

The problem is that organism would be massively too expensive to survive.

OpenAI's are worse and the market thinks so; they are having trouble raising the money they want, the way they want.

Agree with the rest of this though. This whole idea that Nvidia is likely to get much of a normal return from their vendor financing strikes me as tenuous, and what they really are getting is the creation of the perception of scarcity.

One day soon, the notion that GPU scarcity is a proxy for AI demand will finally lose its currency and I think that will be the start of the reckoning.

It’s not the digital world where you just duplicate an object in memory. You need to find the power somewhere (new contract, generator installation), route it, add management tools (and maybe technicians), and then plug the new gpus. It’s all physical work, which also does cost money, take time, and maybe constrained.

ah this is the category error of the whole bubble.

We are so use to the digital world and the efficiency of electronic markets that it is hard to think of physical constraints in the real world.

> Do you truly believe that the cost that can be charged for inference is below the cost of electricity?

That isn't the issue - it's that building brand new datacentres + the power generation to run them takes a massive amount of time/permitting/money, and by the time those datacentres actually come online, you can be sure that the price pressure on inference will be sharply negative.

Nvidia does lease them back from some of these customers. But it's not clear even Nvidia is plugging them all in.

Car companies do that all the time ever since banking licenses have been issued to non-financial institutions.

Nothing new, nobody pays a car cash anymore

I wonder if the economics of the two industries is similar enough to be useful here. Car demand itself seems fairly stable and repossession is a way of recovering value in cases of default. Credit scores of individuals purchasing cars being tied to salary/wages vs. startups without much history. I just wonder if the analogy isn't fraught with issues.

Most of the GPUs that have been sold haven't been plugged in because there isn't enough electricity in the world

replace world with US

there is plenty of electricity in china

There's plenty. Electricity (and energy) is absurdly cheap in the modern world. 1GJ is roughly 10$. An H100 running 24/7 takes 20GJ/year, so that's 200$, which is orders of magnitude less than its retail price.

If energy wasn't cheap, you wouldn't see people driving large vanity SUVs around in the least economic way possible.

Something something something pride before a fall something.

It’s very much not true in all geographies.

It also requires major infrastructure to deliver that energy, which takes money and time to build, as do the datacenters.

"If cake weren't cheap, me and the other aristocrats wouldn't eat it daily"

Around 0.05-0.15% of the world population own such large SUVs.

Just FYI, there are around 360 mil SUV's on the road today around the world. Let's say 1/3 are "large", 120 mil. So, out of population of 8.3 bil, that's 1.4% or apx 3% of households. :-)

You mean there are places and times when electricity is cheap and available, and it if would last 24/365, then it would take 200$ to run such a card. The thing is: it is not available 24/356 in that quantity at that price.

Can I get a source on that please? I think the talk of not having enough power for the AI rollout refers to future needs not current demand.

There's plenty of electricity to run every GPU in the world, just maybe not enough electricity to run every GPU in the world and have poor people afford AC.

Great! How steep a discount can I get to take them off the hands of people who can't plug them in?

Right? This fever dream that there are boxes upon boxes of unused cards (some even claiming on this thread that it is 50% of GPUs sold) is utterly delusional. Some people want AI to go away so badly that they temporarily lose the ability to reason.

You're giving them a lot of leniency. Personally I get the impression many never had the ability to think in the first place, just imitate it by regurgitating what they've heard from others.

I try and give a lot of leniency because generally these people aren't stupid or meaning to be unpleasant, they are just scared. Which means sympathy, kindness and understanding is the likely cure. You can't reason someone out of a position they didn't reason themselves into.

Because most hyperscaler projects are stalling. The number of datacenter chips number was way higher in Zitron's report.

While your datacenter is not built, guess where the AI chips are? That's correct. In boxes.

> While your datacenter is not built, guess where the AI chips are? That's correct. In boxes.

Are you sure about that? Might just as well be in Excel sheets only...

I'm willing to believe Amazon is sitting on bunch of unused GPU[1].

[1]https://www.techspot.com/news/114315-dark-gpus-amazon-show-l...

With the recent craziness, perhaps not all of it is as far from the truth as we might think? There could be quite a gap between units "sold" and units actually made...

If it were to be worse than channel stuffing — that is to say, if Nvidia has sold GPUs that exist only on paper — that would be beyond bad.

FWIW we absolutely know for sure that both AWS and Microsoft have racks upon racks of GPUs that are unused, if the cards are installed at all.

(AWS could have $20B of unused GPUs at this point — it's widely reported that there was at least $5B in August, with up to $3B more arriving each month)

https://www.businessinsider.com/amazon-nvidia-ai-gpu-data-ce...

We know that Satya Nadella has been open about how one of his big problems (as a CEO) is that he has bought GPUs he cannot plug in anywhere.

So 50%? Who knows. But if AWS and Microsoft can't plug them all in, why assume anyone else can?

The clear indication is that cold/dark GPUs delivered to major customers, GPUs that Nvidia has sold but still has on its shelves (buy and hold), and GPUs that Nvidia has sold, leased back from customers and may or may not have plugged into their own cloud, together represent a large enough proportion of sales to have pushed up the prices of GPUs and caused scarcity that is somewhat artificial.

If you take the scarcity of GPUs to be a solid proxy for assessing compute demand — or even for the sales of GPUs to be a solid proxy — then you are making a dangerous bet, IMO.

Step 1: wait for the financial bubble to pop

Step 2: still have money even though the financial bubble popped

Step 3: just buy them normally, at market prices

This is a preposterous statement to make.

When a bubble pops, the pop typically takes years to happen. It's more of a slow deflation in most cases. Afterwards, you can easily spot the peak that would never be reached again but, during a "pop", it's not clear if you're past that peak or if you're just experiencing a momentary blip on the path to new highs.

Nvidia is likely not going to surge back to $240 soon. It might go lower actually. Skeletons are still falling out of closets. Huang may have been the tech founder's tech founder, but he kept some pretty nasty company in recent years and, as we're now finding out, got both himself and his company into shady territory.

Is Nvidia slowing down for a year or two bad for AI as a technology? Probably not. Nvidia could find a leader that will return them to top form, or another company could take advantage of the situation to catch up and surge past. A lot of AI researchers have been calling for a pause or, at the very least, a slow-down. This might be just what was needed.

However, given Nvidia's utterly unique position as both hardware supplier and bank for the data centre build-out, what will the impact be on the rest of the AI sector? The market is going to have to price in a weaker and slower build-out as well as a loss of trust in Nvidia. That's clear downwards pressure. Is it a blip, or are we now past the peak?

It’s not the only indication that we are past the peak, I think. OpenAI’s failure to raise what they wanted, Anthropic’s likely junk rating, banks struggling to offload AI debt at fairly deep discounts, Oracle’s obvious problems.

As you say, if Nvidia is chastened enough to slow its banking endeavours and that lets air out of the bubble in a progressive rather than dramatic way, that’s probably better for everyone (except OpenAI, who are probably hosed in that scenario, as they have huge commitments but are struggling to raise what they need; Anthropic might last long enough to reassess)

I'll admit it, I am not very well educated in the ways of finance but I am a bit baffled how Anthropic can be valued at (reportedly) upwards of 1 trillion but also have a "junk" rating?

A lot of people want to buy their stock because of their fame, but the company’s underlying financials are junk (way too many commitments, way too little revenue)

The problem is they don’t have a unique enough product to sell at so high margin. If they were the only AI good enough to use, then I’d be ok. But given that cheap Deepseek is always catching up their stock is useless

I’m also not an expert, at all.

Junk rating from a relatively high likelihood that they can’t pay the people they owe money to. If you buy their shares they are much more likely to go to zero than a more stable company in the near term. So you don’t want to buy anything Anthropic with money that you may need for anything.

High valuation because the public believe that if they do figure it out they could be worth hundreds of trillions. So a lot of people may want to buy it as a speculative bet of ungodly returns

Junk is just a terrible name in 2026.

Previous bubble villain Michael Milken and the 80s leveraged buyout bubble made it a name of common parlance because that bubble was built on "junk" bonds.

The old traders of that time who were against what Milken was doing called them "junk" and the name has stuck ever since.

It just means "risky". If you trade bonds it has no association with "bad". It is an inside bond market term, I think that is why it sticks too. It is a fun term if you trade bonds but terrible for communicating to the average person.

No, junk means something specific here, it's not just a vague term.

If you think it's terrible for communicating to the average person (and I am very much the average person) you could explain it yourself in detail.

So I will try:

You are right that it is a sort of consensus rating threshold maintained by the ratings agencies that means their (Anthropic's) debts are not considered investment-grade — that they are speculative, high risk, that they are considered likely not to be able to repay. It is ultimately the ratings' companies (and therefore some individual people's) decisions as to whether they are or are not junk-rated.

But for Anthropic it also means they will pay much more for their debt, they may not be able to negotiate long repayment terms, they may need to de-risk with underwriters by paying higher fees, they may be limited by the terms of any individual loan — it may limit how much future debt they are allowed to take on, may mean they need the lender's approval to restructure their business in even small ways, etc., and most importantly, some institutional investors (pension funds, mutuals, insurance companies, ETFs etc.) will be barred from investing (which drives why their debt will cost more).

Frankly I think it is baffling that Anthropic are still headed to an IPO when this is what it could look like on the other side.

Bond ratings are (supposed to be; see 2008 for failure case) based on actual, hard financial metrics. Stock valuations are that, plus fortune telling, plus feels, plus Jim Cramer hitting a big red button.

'junk rating' always means bonds, and it's the share price/market cap that's that high.

Very loosely a 'high' (price over earnings) share price is about growth and exciting prospects; high or well rated bond is about stability. (Not to suggest they move linearly in opposite directions though!)

They are different things, the valuation applies to their equity value (reported) but the junk rating is for its debt and how creditors asses the quality of the debt. If a company is burning tens of billions of dollars and is only "operationally" profitable, you would kind of have some doubts about their ability to repay. Those doubts can stand at odds with how the equity market sees the long-term potential of the company.

Would it not be extremely foolhardy for a new CEO to start slow rolling things considering the dominant position Nvidia currently finds itself in? Why would they do that? You say yourself it would clearly be a managed decline in that case. Why would an incoming CEO want that? Don't forget their pay would likely be performance related and probably not tied to decreasing the stock price

I’m pretty naive about this kind of thing, so bear with me, but if your stock is in free fall and you need a new CEO, isn’t there merit in a stabilizing force? No one is ever compensated just for, at least in the near term, stopping the bleeding?

Well the stock is high because shareholders have been predicting massive future growth. If a CEO is going to stabilize the price, that means they have to give investors confidence in the same or at at least similar growth trajectory. Which was aggressive.

Nice analogy but there’s just one problem: it assumes arriving at mars is the end goal. But that’s not how companies seem to operate. They’re about indefinite growth. Which clearly isn’t possible and ends up killing a lot of companies as they eat themselves from the inside just to keep greedy shareholders happy.

There used to be a time when people would invest over the course of decades and enjoy the dividends they received. But these days it’s all just about the short term rewards.

Tragedy of the stock market: infinity can hit a company even if both management and shareholders are fully committed to modest long term stability (apparently this is absolutely not what happened at nvidia).

A company can get into the sights of "infinity believers" even if they try not to. Infinityists will honestly believe that the stock is a very good buy and whenever a "stability lover" has to sell, they will outbid the next stabilityist. Price goes up, it becomes more visible to infinityists and so on. Eventually infinityist will outnumber stabilityist and management will enter "or die trying" mode on promises they never made. Or be replaced.

If they discover tricky shenanigans, they might well announce them publicly to cover their own personal position. At the start, they're not going to have too much invested.

Yes, you publicly announce everything now so you can throw it all at the feet of the ousted leader. If not, enough time will pass where blaming the ousted leader looks like excuses for the current one and just looks bad to the public. Look at the current US administration as an example of how the constant blaming of the previous administration played to the public opinion.

Right. One of the weird side effects of the Jensen Huang report hierarchy is that instead of a handful of direct reports that have stories they need to CYA on, there are over sixty of them! This could make the congressional testimony aspect pretty intense.

And each has plausible deniability of their personal involvement, since the responsibility is so diffuse below Huang.

That’s a good point. A new CEO never wants a slow unwinding, they go for quick pain to set the baseline of their bonus assessment into something workable.

The peak on the semiconductor ETFs was in spring. Even with the bounce semiconductors are 30% down from the peak.

OpenAI not being able to IPO I think is the huge indicator that we are in denial of.

The tide is starting to go out and we are starting to see who is swimming naked.

Jensen and the Jensenverse as the villain of the bubble is the prestige of this bubble's magic trick. That is a show stopper.

OpenaAI as Myspace I think is the second act, yet to come.

[delayed]

But public perception has huge repercussions in business. If screwing Chinese models contributes to bad public perception, business will realign and oust CEOs who do.

> screwing Chinese models

It's funny how the problem of Chinese models ended up affecting Nvidia and the US AI industry in general. Just not in the way it was predicted.

Bu dum da tiss! - Underrated!

You can make a profit while still being ethical. It isn't even hard to do so.

Sure you can, if you are your own one-man business. Otherwise it's exploitation—you're stealing from your workers. This is not remotely controversial. (or profit sharing etc, which is so rare as to mostly not exist)

Stealing from your workers - have you ever been in business in your life? A workman is worthy of his hire - which means you pay your employees what they are worth and keep the excess. If they want to capture all of it they can start their own business.

Only in a zero sum game, which capitalism fairly definitionally is not.

How do you figure? There are certainly places where we manage to create value, but this has little to do with capitalism and more to do with hauling shit out of the earth that we didn't have before.

> When a bubble pops, the pop typically takes years to happen. It's more of a slow deflation in most cases.

That’s just a bear market. Bubbles pop, otherwise we wouldn’t call them bubbles. If you call every downturn a bubble it ceases to be meaningful, and we’d have to come up with a better term for actual bubbles.

Even if you know the timing of the collapse of a cataclysmic bubble what do you do?

All of the investable assets will collapse at the same time. Inflation will evaporate cash. Bonds will worth nothing (if the us gov bothers to repay them).

Shorting. Put options and futures.

That assumes the counterpart is liquid.

Fourth law of thermodynamics: liquidity will be provided by the central bank.

If you know the timing, CFDs it is.

You need to know with ridiculous precision, though. And the triggering event for the collapse usually isn’t predictable. For instance, if we’re seeing the beginning of the end now, we’ll, no-one could have really predicted the current sequence of events, unless they knew about Huang’s misconduct _and_ when it would be leaked.

It is also so lucrative to participate in a bubble that if you focus too much on the problems you will miss out.

I think part of a bubble is knowing the problems but participating anyway. Over time then the problems start to feel normal and while things are good you rationalize why maybe the problems aren't such a big deal after all.