Which means they’re in a bubble because Nvidia’s total assets (85B$) value is less than half of Intel’s (205B$). I refuse to believe that the “potential for growth” of Nvidia is worth anywhere close to 120B$ in actual value even in the next 5 years. I see only two things here: either Intel is undervalued, or Nvidia is overvalued. I think it’s both. When that bubble bursts it’s going to hit very hard for a lot of people because it’s the same thing as the other big tech companies (apple google meta etc) are all valued based on predictions and magic when the companies that have an actual intrinsic value are worth less
Seems the strategy is clear mega short nvidia to buy super leveraged intel and hope you can stay solvent longer than the market is irrationnal
I agree this is an obvious market failure because finance bros have become detached from reality even more than usual.
AI boom is starting to smell like DOT COM 2.0.
We have not seen that much improvement since gpt4, mostly cost reduction and UI convenience.
Current AI hype is not cashable, and I say that as an enthusiast who is building 15 kilowatt inference cluster in his living room.
We already have the major improvements already and we are nowhere close to done disgesting then.
Intel is lagging behind AMD and NVidia with no sign of catching up. Meanwhile NVIDIA has a monopoly on AI.
It’s no wonder NVIDIA is worth far more now.
Intel have shown signs of catching up by putting out a better iGPU than AMD’s latest and greatest for laptop chips in certain games and most compute tasks. They’ve also put out one of the best laptop chips last month, they consume next to nothing while still having decent performance but go on I guess
Aren’t the best handhelds using AMD iGPU’s? The MSI Claw didn’t exactly leave a great mark. The new Radeon 890m looks pretty killer for its power efficiency.
The cpus I’m talking about have released about a month ago, but until very recently AMD were the only good options for handhelds
How is Intel’s market cap less than their total assets?
Debt, both on-the-books and anticipated.
Intel’s investments in the Titanium chipset have effectively dead-ended. They can’t get below 7nm efficiently. Meanwhile, you’ve got companies in Taiwan, Korea, China, and Japan breaking into the 3nm and 2nm scales. To catch up, they’re looking at hundreds of billions if not trillions of dollars in technical debt.
Yes, they can keep churning out existing processors at huge profits in the moment. But the face value of these processors plummets with every new step in Moore’s Law. This amounts to asset depreciation, which means Intel’s value is heavily overstated on the basis of asset cost alone.
I won’t argue that NVIDIA is overvalued. But I think the degree to which they are overvalued is often misattributed to speculation and avoids the real specter haunting the company… competition. NVIDIA’s market dominance and the escalating demand for their technology means the sky really should be the limit for their growth. Demand for AI processing is at the forefront of these expectations. But a rival manufacturer capable of cutting into demand for their units would dramatically undermine their profitability.
Its the same with firms like Microsoft and Facebook and Boeing. So much of their dominance is predicated on the theory that people will never leave these walled gardens and their margins being enormous purely because they controlled a critical commodity/patch of technical real estate.
There was - incidentally - another enormous company that seemed to have the market cornered in its industry and got complacent with its R&D and long-term investment strategies… Intel.
Right. Liabilities. Frick. Thank you.
Well, while outstanding in current (we are living in a 20++ year bubble without much correction - no, when market values/indexes/cap rebound within a year, or month, with mostly the same main players) times, and especially for a successful company in the tech sector, having more assets than market cap isn’t that weird.
As long as market value is above book value, it’s fine.
And when it’s not, it’s prob a bank after 2012 (tho prices are generally way closer to book than they were before) :D.Fear mongering articles making it seem like they are doomed and will go bankrupt after bad quarterly results were announced. Articles were probably sponsored by rich people wanting to buy Intel stock for cheap. But they won’t go bankrupt because the US gov./army need Intel to stay relevant against China, and Intel is basically the only American company that both designs and fabs their own processors and that is still relevant.
That and the fact that Nvidia is over valued (they are valued at 30x the value of their assets).
thirty times???
Yup. I wager the vast majority of that is AI hype. Nvidia is the king there and in datacenter GPU compute in general, and investors are betting that Nvidia will continue to dominate and that market will continue to be relevant and grow.
I have my doubts, but as a famous economist once said:
Markets can remain irrational a lot longer than you and I can remain solvent.
So I’m not putting my money where my mouth is just yet.
That is an excelent quote.
Crazy how quickly NVIDIA went up. I wonder if they’ll crash down just as fast should the AI hype either die off or shift to other manufacturers (Intel, AMD etc.) or in-house solutions (ex. Apple Intelligence).
I just want to get a graphics card for less than the rest of a rig combined… shits ridiculous, and AMD doesn’t seem to be even trying to compete anymore
they do compete, its just users weigh DLSS and Raytracing far more than they should, and devalue VRam in long term situations
for example a 7900 GRE cost about the same as a 4070, but more people will buy the 4070 regardless
I definitely do like raytracing, sadly. I’m more interested in graphics and immersion in a setting/story in a game than competitiveness or ultra-high FPS. Water reflections and mirrors just look absolutely gorgeous to me.
I’m definitely strongly considering AMD regardless for my next build, as I’d like to switch to Linux fully at some point.
Eh, I got an AMD GPU somewhat recently and it meets all my expectations. I’m not too interested in RTX or compute, and they have a really good value on raster performance.
I’m coping for RDNA4.
Apple is not there yet, its models were trained on Google hardware. Though I am surprised it wasn’t Nvidia hardware.
What’s “Google hardware”? Likely just NVIDIA hardware running in Google’s cloud?
No no, Google does actually have its own custom proprietary AI hardware - https://en.wikipedia.org/wiki/Tensor_processing_unit
Ah, TIL.
This is all so normal and sustainable.
Yes, go on, let NVIDIA buy intel. Let them buy AMD too. What could go wrong. I love monopolies! /s
I wonder how this guy feels when he watches everyone in the streets trying to make ends meet.
Probably feels he deserves it. :)
He’s being doing nothing but selling his Nvidia stock so I’m pretty sure he knows it is overvalued
If this is all Nvidia stock let him try to cash out and see what happens.
In case anyone was curious - the avg daily turnover of nvda is 300 to 350 million moneys.
The thing that bothers me when people say “oh its unrealized gains, it’s not real money” is that they use those unrealized gains as collateral for loans of real money. They effectively ARE that rich.
It’s BS that you can borrow against it. If he did sell it the valuation would drop.
Banks don’t take this into consideration when assessing collateral?
Lol, who downvotes a question?
There are completely different rules when you are that rich. Look at Trump, he bankrupted how many businesses and banks STILL lined up to loan him money. At the very top, your trading favors and power.
Take what into account? They basically look at current valuations and offer loans up to some fraction of that amount.
And that’s generally the way the ultra-rich operate, they don’t actually sell anything, they just borrow against their assets. They punt the can down the road until they die, at which point those unrealized gains get stepped up in basis for those who inherit it. If you have enough stock assets, you can service the debt with the capital gains you’re forced to realize (i.e. dividends).
So the bank sees someone with $100B in assets asking for a $10M loan or whatever, and they’re completely happy to offer that, because even if the stock gets cut in half, he can still pay the debt.
So how do they pay the loans back if the liquidity isn’t there?
With money they loan from a bank, using whatever they bought with the previous loan as collateral.
It’s credit all the way down.And that ends when they die, at which point the stocks get stepped up in basis so the taxes are almost completely avoided. Or they structure their debts in such a way that certain entities can be bankrupted without impacting the actual assets.
Things get wild when you’re in the 0.1% and above.
You dont need to sell your stocks to access that wealth. You can use that as collateral to take loans or exchange stocks.
“unrealized gains” that you can somehow live off of indefinitely.
Elons everything comes from having overpriced tesla stock as collateral
LOL
“AI” technologies (in case of complex “universal” models, and not ML for narrow applications) suck for the same reason oil product based technologies suck. And that’s not global warming and microplastics. That’s power centralization and globalization the wrong way.
Both those enormous datasets and dedicated hardware for them are points of centralization.
No, he’s not. It’s only his inflated market value.
He can turn a significant chunk of this value into actual dollars, even without selling the stock. This line of reasoning that execs’ worth is not what it seems to be because it’s based on share value is constantly used to discount their wealth and argue against acting on wealth inequality.
Exactly. At the very least, he could go get a margin loan at relatively low cost (like 5%) compared to the tax burden of cashing out (20% or more). And that’s just using publicly available numbers, a billionaire can get a lot cheaper loans than that.
Both numbers are based off market value, though
