That's exactly the thing. It's only about bookkeeping.
The big AI corps keep pushing depreciation for GPUs into the future, no matter how long the hardware is actually useful. Some of them are now at 6 years. But GPUs are advancing fast, and new hardware brings more flops per watt, so there's a strong incentive to switch to the latest chips. Also, they run 24/7 at 100% capacity, so after only 1.5 years, a fair share of the chips is already toast. How much hardware do they have in their books that's actually not useful anymore? Noone knows! Slower depreciation means more profit right now (for those companies that actually make profit, like MS or Meta), but it's just kicking the can down the road. Eventually, all these investments have to get out of the books, and that's where it will eat their profits. In 2024, the big AI corps invested about $1 trillion in AI hardware, next year is expected to be $2 trillion. Only the interest payments for that are crazy. And all of this comes on top of the fact that none of the these companies actually make any profit at all with AI. (Except Nvidia of course) There's just no way this will pan out.
There are three distinct but related topics here, it's not "just about bookkeeping" (though Michael Burry may be specifically pointing to the bookkeeping being misquoted):
1. Financial depreciation - accounting principals typically follow the useful life of the capital asset (simply put, if an airplane typically gets used for 30 years, they'll split the cost of purchasing an airplane across 30 years equally on their books). Getting this right has more to do with how future purchases get financed due to how the bookkeepers show profitability, balance sheets, etc.. Cashflow is ultimately what might create an insolvent company.
2. Useful life - per number 1 above - this is the estimated and actual life of the asset. So if the airplane actually is used over 35 years, not 30, it's actual useful life is 35 years. This is to your point of "some of them are 6 years old". Here is where this is going to get super tricky with GPUs. We (a) don't actually know what the useful life is or is going to be (hence Michael Burry's question) for these GPUs (b) the cost of this is going to get complicated fast. Let's say (I'm making these up) GPU X2000 is 2x the performance of GPU X1000 and your whole data center is full of GPU X1000. Do you replace all of those GPUs to increase throughput?
3. Support & maintenance - this is what actually gets supported by the vendor. There doesn't seem to be any public info about the Nvidia GPUs but typically these are 3-5 years (usually tied to the useful life) and often can be extended. Again, this is going to get super complicated to financially because we don't know what future advancements might happen to performance improvements to GPUs (and therefore would necessitate replacing old ones and therefore creating renewed maintenance contracts).
Typical load management that’s existed for 70 years: when interactive workloads are off-peak, you do batch processing. For OpenAI that’s anything from LLM evaluation of the days’ conversations to user profile updates.
Flops per watt is relevant for a new data center build-out where you're bottlenecked on electricity, but I'm not sure it matters so much for existing data centers. Electricity is such as small percentage of total cost of ownership. The marginal cost of running a 5 year old GPU for 2 more years is small. The husk of a data center is cheap. It's the cooling, power delivery equipment, networking, GPUs etc that costs money, and when you retrofit data centers for the latest and greatest GPUs you have to throw away a lot of good equipment. Makes more sense to build new data centers as long as inference demand doesn't level off.
Emmentaler and Gruyère close in texture and flavour and interchangeable? Oh boy, if you ever travel to Switzerland, I recommend you to keep your opinions on cheese to yourself. ;)
Emmentaler has holes, is low in salt and has a taste which is on the bland side (I personally don't like it), whereas Gruyère has no holes, is saltier and has a much richer and "rounder" flavour. It comes in different stages of ripening, from young, which is soft in texture and mild, to old, which is hard and has a much stronger flavour. I personally like Gruyère mi-salé a lot, the semi-ripened one. It's close to a perfect cheese if you ask me.
TIL that Gruyere from France is different to Swiss and it must have holes.
https://en.wikipedia.org/wiki/Gruy%C3%A8re_cheese
> The PGI documentation also requires that French Gruyère has holes "ranging in size from that of a pea to a cherry", a significant departure from the Swiss original. Peter Ungphakorn, a Swiss local and an international trade expert, comments that the French Comté cheese could be a closer match to the Swiss version.
As a Swiss, I can assure you that this is false. Most cheese varieties have very strict quality requirements, if they're not met, the cheese may only end up as no-name ground cheese for pizza or something like that. But an Emmentaler, Gruyere, or Sbrinz always has the same quality, no matter if it's exported or for domestic consumption.
As a Frenchman I disagree, cheese is very sensitive to environmental condition, in particular during transport. To eat a good piece of St Nectaire, first go to to St Nectaire (eat the crust too!)
The confusion between force and mass tends not to get appreciated by many folks outside of engineering. Particularly in aerospace when you slip the surly bonds of earth and your mass stops being so bound to the force your feet impart on the floor.
I can hear the start-up of 777 engines at the airport, every night at approx. the same time. The airport is 12 km from my apartment and there's a hill in between. But my bedroom is facing the airport side and the wind mostly comes from that direction. It's crazy.
According to the article, at least todays retinal-based photosynthesis is anoxygenic and does not invole carbon fixation. At night, these cells metabolism stops. Chlorophyllic photosynthesis with attached carbon fixation allows the cell to build up starch during the day, which it breathes under the use of oxygen at night, so the cell remains active during the night. Looks like a big evolutionary advantage to me.
Also, light is not the limiting factor for plant growth, it‘s usually water or nutrient availability.
carbon fixation is a completely separate process. in principle you could hook up a sufficiently engineered cell to electrodes and do the carbon fixation part in the dark by supplying it with juice from the mains.
accordingly there is no particular reason for purple photon assimilation to not be attached to carbon fixation... though i suppose as the electron energy levels dont quite match up it might be a schlep to get purples to make sugar.
In other news, Proton VPN and probably email are also subject to surveillance by the Swiss Federal Intelligence Agency NDB, including decryption of VPN traffic and data retention for up to 6 months. Switzerland is no save haven for your communication needs.
reply