No it's cooked. For high tech items, they assume that improved technology means you are getting more for your money even if the price goes up, so they discount it. It's true that you get more for your money, but it ignores threshold effects, like you just can't buy an equivalent phone for $10 even if todays phone's are 200x better.
Then there's the "owner's equivalent rent" BS and this is 25% of CPI. It answers the question "If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished, and without utilities?" It assumes rental price and housing costs are somehow linked when in reality asset prices have far outstripped rent.
> it assumes rental price and housing costs are somehow linked when in reality asset prices have far outstripped rent
It's pricing the cost of shelter. Renting a home is buying shelther. Buying a home is buying shelter and buying a financial asset. OER is the way you separate the last two components. Otherwise, you'd have to only look at rents to determine housing prices, which would be rubbish in a country where most households live in homes they own.
Suppose you want shelter. It used to be that you could buy a house for a reasonable price and move in. But now that's unaffordable, so instead your have to pay rent. In CPI thinking these are equivalent forms of shelter, but I bet if you asked most Americans, they would not agree with you.
Which flows through to owner-equivalent rent, in part.
> In CPI thinking these are equivalent forms of shelter, but I bet if you asked most Americans, they would not agree with you
It really doesn't. When measuring rent, you directly measure rent. For OER, you're measuring the housing price and imputing shelter cost from that. They're similar, but different. Sort of like how renting and owning are similar, but different.
Also, given the variance in housing affordability across the country, you'd almost certainly have to strip out any financial-asset component anyway to meaningfully compare the resulting number.
why should the asset prices matter in OER? the aim is understanding cost. BLS no longer questions homeowners but samples local rents to estimate OER since homeowners could've been wrong in their guess. of course, someone may have locked in a low interest rate so their expense is overstated. counterargument is that you are consuming a more valuable service by occupying the unit even though market rent exceeds your costs so it doesn't matter if your cost is assumed to be the market rent. note there is a 6-month sampling lag of rents, which doesn't help the perception gap in the inflation figures.
This might not be true. Someone was comparing Nvidia's production rate with known data center capacity, and they do not match. Their conclusion was that people (possibly even Nvidia) were hoarding GPUs- in the very short term this might be a good strategy, but GPUs go EOL fast. There are other stories about paused datacenter builds that match with this.
TSMC is definitely fully allocated, based on current 40 wk lead times for FPGAs..
All that means is that there's a bottleneck at the data center layer. When he says "dark GPUs" he's saying that there are no dark DEPLOYED GPUs.
This is a reference to the 1990's dot com bubble where internet infrastructure companies overbuilt network capacity, leading to the term "dark fiber". That was an indicator of a bubble because it showed that capacity was larger than demand. OP is saying that this is specifically NOT happening in the case of GPUs yet, indicating that demand still outstrips supply of compute.
>GPUs go EOL fast
We are seeing the opposite of what was expected, GPUs are actually getting more valuable because demand is so great, something that basically never happens. Even older chips have become more valuable.
>paused datacenter builds
It doesn't seem that datacenters have been paused because of lack of demand for AI, it seems mostly that there is a lot of pushback by cities to build these things and also there is a shortage of power to run them.
IMO none of these things point to a AI being a bubble (over-hyped, demand does not match the stated value). It mostly points to the opposite, there is massive demand for AI and every layer of the supply chain is struggling to keep up with that demand.
Adding to this, a lot of fiber installed in the 1990s is still dark. Multi-wavelength XYZ and other improvements mean the same fiber from 35 years ago can carry 100 or 1000x what it was originally designed for. Also, like Solar, all the cost is in labor. When they designed the Seattle/King County fiber network, they knew they would never have access/permits to go back and add more, so instead of running a single 12 fiber bundle the size of your pinkie, they ran 3 x 1024 bundles the size of your arm through the hollow bridges that span I-5 freeway and snakes through Seattle underground. Almost all of that sits dark today despite being in a very busy area, simply because fiber technology keeps getting better.
Yea, fiber is great. They can do hundreds of terabits/s per fiber today, and petabits/s is not far away. Bandwidth is fundamentally cheap enough that my ISP offers 50Gbps residential service!
I live in Oregon. The price was $900/month last time I checked. I believe they do provide a QSFP+ module to plug into your equipment. They also allow residential customers, at any tier of service, to announce their own IP blocks via BGP.
> ...GPUs are actually getting more valuable because demand is so great, something that basically never happens. Even older chips have become more valuable.
Huh, anybody want to buy a GTX 680? Or even a formerly-SLI'd pair?
Don't you think that under excess demand, production will ramp, competition will become available etc? These posts read like we're all out of fresh silicon or something.
Supply will catch up, it will just take 3-5 years, with the price rising the whole time. Basically a worse version of the Covid supply disruption where I sold my car for more than I bought it for years later.
The physical world can’t be patched overnight, and cutting edge manufacturing takes a long time. Fortunately we are in a very peaceful low tension world right now and no one would try burning down or blowing up one of those extremely important, irreplaceable fabs.
> IMO none of these things point to a AI being a bubble (over-hyped, demand does not match the stated value). It mostly points to the opposite, there is massive demand for AI and every layer of the supply chain is struggling to keep up with that demand.
Yes, the demand is there for the currently unsustainable price. Lets see what happens when the dumping of money into AI stops and the companies are forced to increase prices a lot.
> IMO none of these things point to a AI being a bubble (over-hyped, demand does not match the stated value).
I agree the demand is there, but hyperscaler capex is what now? 3% GDP? This is an absurd amount of money and people who question whether the ROI is there have a point just because of the order of magnitude of this spend number.
Yeah, it's how you pay dividends at capital gains rate instead of income rate. But also, Apple is kind of stagnant.. if they had something better to do with the money they wouldn't be doing the buybacks.
Not only that, the operating itself was configured by a process known as "sysgen". You relink the entire operating system with options set the way you wanted. It was generally a miserable slow process.
At least on PDP-11 that's how it went for something like RSX-11. I believe same is true for early IBM mainframe operating systems like DOS 360- I think all programs had to be relinked because one option you had was to move things around in memory, and ancillary programs had to know the memory map.
Even later: I wrote a device driver for Xenix: you got a link kit for the OS, you relinked it with your custom driver object file included.
On CP/M you patched the running image (perhaps with DDT), then use the CP/M SYSGEN command to install it on a disk.
CP/M 1.0 was demoed to Intel in 1974, but they didn't buy it.
iCOM FDOS was the first operating system that was available to people, and it sure didn't have environment variables.
Anyway, these operating systems didn't have multiple directories. But you could use CP/M 2.x's ASSIGN command to bind a logical name to a physical name. Minicomputer operating systems had this, also IBM mainframe had JCL DD commands.
Includes GCD and LCM, some of the newer ones don't have them.
If you want graphing, there is the newish fx-CG100 has a nice display, but they removed Casio basic, it now only has micro Python (way too awkward to type on a tiny keypad):
I agree with you on the Casio fx-115ES Plus 2nd Edition. I picked one up two years ago for $11.41. It naturally writes out equations, has a backspace and is generally excellent. I still love my HP RPN calculators, but the fx-115ES works nicely for anyone who isn't using RPN or sympy.
Then there's the "owner's equivalent rent" BS and this is 25% of CPI. It answers the question "If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished, and without utilities?" It assumes rental price and housing costs are somehow linked when in reality asset prices have far outstripped rent.
reply