Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Building Meta's GenAI infrastructure

engineering.fb.com

Post reply on HN
Print view View original

291–300 of 314 posts

Re: Building Meta's GenAI infrastructure

#291

Post by woooooo »

whatshisface wrote: post #241
boringg wrote: post #240

Earlier quoted context omitted.

Front loading costs to eventually extract rents on usage with one hell of a capital wall protecting the assets. Its easier to spin up a business for sure -- also easier to unwind it - there not as sticky as they used to be.

If the government can stay back far enough that more than one AI company can train their models, it will end up working like steel mills - barely enough profit to pay the massive cost of capital due to competition. If the government regulates the industry into a monopoly, all bets are off. Their investors are going to push hard for shutting the door behind them so watch out. The only question is - what tactic? I don'…

Funny example, US Steel was a textbook case for a monopoly achieved privately because it wasn't regulated against.

Re: Building Meta's GenAI infrastructure

#292

Post by oblio »

elbear wrote: post #262
krschultz wrote: post #259

Earlier quoted context omitted.

Because they make more money using their servers for their own products than they would renting them to other people. Meta has an operating margin of 41% AFTER they burn a ton on Reality Labs, while AWS has a 21% margin with more disciplined spending. Social media is a more profitable business than infrastructure.

Does Meta make money from anything other than ads? It's not a dismissive question. I'm curious if social media implies anything other than ads.

> Advertising (over 97.8% of revenues): the company generated over $131 billion in advertising, primarily consisting of displaying ad products on Facebook, Instagram, Messenger, and third-party.

https://fourweekmba.com/how-does-facebook-make-money/

Re: Building Meta's GenAI infrastructure

#293

Post by the-rc »

virtuallynathan wrote: post #284
the-rc wrote: post #222

Earlier quoted context omitted.

Mapping as in.. drawing the outlines of buildings and computing the square footage yourself?

Yep.

Then you should be aware that, for the longest time, Google was against multiple floors, until they suddenly switched to four floors in many locations:

https://www.datacenterfrontier.com/cloud/article/11431213/sc...

A decade ago, there was a burst in construction and in some places the bottleneck was not getting the machines or electricity, but how fast they could deliver and pour cement, even working overnight.

Re: Building Meta's GenAI infrastructure

#294

Post by boringg »

brookst wrote: post #242
boringg wrote: post #240

Earlier quoted context omitted.

Front loading costs to eventually extract rents on usage with one hell of a capital wall protecting the assets. Its easier to spin up a business for sure -- also easier to unwind it - there not as sticky as they used to be.

This is typically called a high fixed cost business, like airlines, hotels/apartments, SpaceX, etc. The dream may be barriers to entry that allow high margins (“rents” if you prefer the prejudicial), but all too often these huge capital costs bankrupt the company and lose money for investors (see: WeWork, Magic Leap). It is high risk, high return. Which seems fair.

I understand the economics concept. I'm not sure WeWork was a great example it had significant other challenges such as a self-dealing founder and, frankly, a poor long term model.

I would wager that the concept needs a bit of a refresh as historically it has referred to high capital costs for the production of a hard good though in this case there is more than just a good produced theres a fair bit of influence and power associated with the good and a ton of downstream businesses that are reliant upon it if it goes according to plan.

Re: Building Meta's GenAI infrastructure

#295

Post by nl »

dougdonohoe wrote: post #191

Having lived through the dot-com era, I find the AI-era slightly dispiriting because of the sheer capital cost of training models. At the start of the dot-com era, anyone could spin up an e-commerce site with relatively little infrastructure costs. Now, it seems, only the hyper-scale companies can build these AI models. Meta, Google, Microsoft, Open-AI, etc.

I too went through the dot com era: as in when Sun Microsystems had the tag line "we are the dot in dot com".

I assure you that before Apache and Linux took over that "dot" in the .com was not cheap!

Fortunately it only really lasted maybe 1993-1997 (I think Oracle announced Linux support in 1997, and that allowed a bunch of companies to start moving off Solaris).

But it wasn't until after the 2001 crash that people started doing sharded MySQL and then NoSQL to scale databases (when you needed it back then!).

It's early. You can do LORA training now on home systems, and for $500 you can rent enough compute to do even more meaningful fine-tuning. Lets see where we are in 5 and 10 years time.

(Provided the doomers don't get LLMs banned of course!)

Re: Building Meta's GenAI infrastructure

#296

Post by virtuallynathan »

the-rc wrote: post #293
virtuallynathan wrote: post #284

Earlier quoted context omitted.

Yep.

Then you should be aware that, for the longest time, Google was against multiple floors, until they suddenly switched to four floors in many locations: https://www.datacenterfrontier.com/cloud/article/11431213/sc... A decade ago, there was a burst in construction and in some places the bottleneck was not getting the machines or electricity, but how fast they could deliver and pour cement, even working overnight.

Yep, I am aware, I have a square footage multiplier for their multi-story buildings.

Re: Building Meta's GenAI infrastructure

#297

Post by renegade-otter »

boringg wrote: post #294
brookst wrote: post #242

Earlier quoted context omitted.

This is typically called a high fixed cost business, like airlines, hotels/apartments, SpaceX, etc. The dream may be barriers to entry that allow high margins (“rents” if you prefer the prejudicial), but all too often these huge capital costs bankrupt the company and lose money for investors (see: WeWork, Magic Leap). It is high risk, high return. Which seems fair.

I understand the economics concept. I'm not sure WeWork was a great example it had significant other challenges such as a self-dealing founder and, frankly, a poor long term model. I would wager that the concept needs a bit of a refresh as historically it has referred to high capital costs for the production of a hard good though in this case there is more than just a good produced theres a fair bit of influence and…

It's more like "disrupting the market". The problem is that it's a whole market.

Uber just now turned its first profit since 2009, and I would wager that if not for the newly found appreciation of efficiency and austerity, it would still be burning through money like a drunken socialist sailor.

Classic approach required basic math. "Here is my investment, here is what I am going to charge for rent". You actually can figure out when your investment starts paying off.

This new "model" requires tall, loud, truth-massaging founders to "charm" VCs into giving away billions, with the promise of trillions, I guess. The founders do talk about conquering the world, like, a lot.

I do not know what the WeWork investors were thinking when they expected standard real estate to "10x" their money while the tenants were drinking free beer on tap. The whole thing screamed "scam" even to a lay-person.

Re: Building Meta's GenAI infrastructure

#298

Post by brookst »

bennyelv wrote: post #258
brookst wrote: post #244

Earlier quoted context omitted.

I’m old enough to remember the proud, defiant declarations that the internet was just a hype cycle.

Well it was wasn’t it? There was a massive boom where loads of companies over promised what they would achieve, followed by a crash when everyone realised lots of them couldn’t, followed by stability for the smaller number that could. It was the very definition of a hype cycle as far as I can see. Hype cycle doesn’t mean “useless and will go away”, you have the second upward curve and then productivity. https://en.m.…

I don’t disagree, but a lot of “analysis” was not that nuanced. At one time I worked for a company where 90% of revenue was from printed periodicals. Smart, capable executives assured the whole company that the internet was not a threat, just something college kids used for fun.

Colloquial, dismissive use of “hype cycle” does not usually mean “this will change the world but foolish things, soon forgotten, will also be done in the short term”. Though I agree a deeper understanding of the term can suggest that.

Re: Building Meta's GenAI infrastructure

#299

Post by brookst »

boringg wrote: post #294
brookst wrote: post #242

Earlier quoted context omitted.

This is typically called a high fixed cost business, like airlines, hotels/apartments, SpaceX, etc. The dream may be barriers to entry that allow high margins (“rents” if you prefer the prejudicial), but all too often these huge capital costs bankrupt the company and lose money for investors (see: WeWork, Magic Leap). It is high risk, high return. Which seems fair.

I understand the economics concept. I'm not sure WeWork was a great example it had significant other challenges such as a self-dealing founder and, frankly, a poor long term model. I would wager that the concept needs a bit of a refresh as historically it has referred to high capital costs for the production of a hard good though in this case there is more than just a good produced theres a fair bit of influence and…

Agreed, and Magic Leap had its own problems. My point was just that “invest huge amounts of capital to create a moat and then monetize in the long run” Is an inherently risky strategy. Business would not work if society insisted that large, high risk investments could not product higher long term margins than less risky investments.

Re: Building Meta's GenAI infrastructure

#300

Post by jedberg »

virtuallynathan wrote: post #296
the-rc wrote: post #293

Earlier quoted context omitted.

Then you should be aware that, for the longest time, Google was against multiple floors, until they suddenly switched to four floors in many locations: https://www.datacenterfrontier.com/cloud/article/11431213/sc... A decade ago, there was a burst in construction and in some places the bottleneck was not getting the machines or electricity, but how fast they could deliver and pour cement, even working overnight.

Yep, I am aware, I have a square footage multiplier for their multi-story buildings.

But how can you know how many floors they have? And where are you getting the list of buildings from? And what makes you think your list is complete?

Also how do you know their efficiency? Google might have less space but also a way to pack twice as much compute in the same place.

Like I said, this is impossible to know without a lot of insider information from a lot of companies.

Post reply on HN