I'm not trying to say it's not an advantageous model for the industry. I just don't think this particular system is scalable, or in fact even that good. It's just an accomplishment for what they had to work with. It's not better than 2 NVL72's strapped together. And in order to match both performance, space, and efficiency, I think they need to start from scratch and design something else. Maybe this advancement will help them create better chips so they can, but I am doubting it.

I'm going to guess, not much further than where they are now. This type of brute force can only go so far. I don't think they have the skills to effectively refine this down.

It's kind of worthy of achievement. But like, it's 5x the number of chips, for only ~2x performance, 3.6x memory capacity, and 2.1x memory bandwidth, while consuming 4.1x the amount of power and taking up at least 4x as much physical space.

Also, though I can't be sure, this type of setup almost begs to suffer from diminishing returns, so I doubt it can be scaled up further

The cheap option probably wouldn't be fiber to begin with. And there are other ways to limit these customers from being too draining. And none of it matters at all if the government subsidizes and pays for the rest, which is already an option. And regardless of what else is required, I'm only paying for digital services. I don't care about phone support. I don't care about online support. I don't care about the technician that might need to come once for setup. For all intents and purposes these are separate things the company offers. You could charge extra to have access to these, or charge based on how much the customer uses them. Or you can bundle it up in a big more expensive package that includes it all. The point is, there are ways to do it without it being a giant drain on the bottom line. Not every single product or service needs to be profitable.

Oversubscription is more nuanced than that, though 32:1 is common. And business users are still higher maintenance than home users. Businesses don't need as much help in the building like home users might need in theirs, but they are much higher maintenance customers for everything else. Business lines need (and pay extra for) maximum uptime. Business plans pay for and expect higher levels of support.

MBAs are garbage degrees. especially if it's the first and only degree you have. Might as well get a degree in circus management.

Correct. I did not know they were separate, because they do not offer fiber at all here. Here they offer 100/10mbps for $55 as their fastest. Technically there are also the wireless ISPs but I don't count them, as I assume you don't either.

Eh, average out the lowest x% of users and what they theoretically cost. I don't think it would be hard if you have access to all the numbers. But also, this would be explicitly for provable low income or elderly type people or something along those lines. People who should have regular and reliable access, but otherwise also likely to be lower on consumption. Yes, it would technically be a "drain" on finances, but if a small enough portion of the customer base, then that can be offset by the more expensive higher end services that offer more. Increasing the profit margin at the top so you can cover the lack of a profit margin at the bottom helps cover a wider spread. And, if the government mandates a $15 option, ideally the government would help compensate the "loses" for each of this class of customer as an incentive to offer it. Not sure if that ever did happen, but that would be the way to do it correctly.

Well, yeah, and because often time price controls are also applied too broadly. However, price controls on physical products and services differ to digital ones. Also, being forced to offer a cheap option, isn't a price control.

Am I claiming that all for-profit ISPs are gouging their customers? Yes. Yes, I am. By how much varies.

Because you are inherently bad with money.

I know this. I just didn't pay close attention to the fact that you cut it off at just private. Brain kept going.

In many other sectors, yes. There are numerous reasons to this and not applicable to this topic.

Because you already have decent competition and no local municipal broadband to compare to (yet). In my area, Comcast or CenturyLink were legitimately the only options, and CenturyLink is absolute trash. Comcast was charging $150+ for gigabit down and garbage up. My city said let's do it ourselves, Comcast complained and spent millions of dollars trying to fight it. Now, suddenly, they have no problem doing the gigabit for $75. Why were they charging $150? Was $150 needed to cover the costs of my plan and usage? No. More than half of it is going to get spent on things completely unrelated to maintaining the service or making it better. They charged more because simply because they could.

They do, but not really because of anything ISP related. They hate Tucows mostly as a domain registrar because of the fact that they had zero problems registering domains that hosted child pornography and letting them continue to run without issue. I don't know about you, but any company that actively allows child pornography to be disseminated is automatically bad and shouldn't be trusted with anything else.

Which I am neither arguing for or against, specifically. They should all offer some sort of minimal, "at cost" priced service though - the exact price should probably be relative to the local area instead of fixed. Either that, or competition has to be forced for areas that don't have any. In my opinion, every single household should be serviceable by 3 different ISPs. And the ISPs have to pay for the costs to hook up these households, even if the household *doesn't* choose them. Is that actually feasible in rural areas? Probably not, but it should be.

However, they aren't responsible for the out of control healthcare costs to begin with. Many times price controls only fail because of the greed of top management. They would rather purposely make the service/product shittier than take a smaller paycheck, so that they can say the price controls are the problem. Most businesses aren't managed particularly well, just well enough.

Private, for-profit companies are inherently designed to offer the lowest possible quality for the highest possible price.

Or if you prefer your money to be spent on the service your paying for instead of a bunch of bullshit like useless (and misleading) advertising and (up)sales people. Municipal internet is faster, cheaper, more reliable, and without datacaps. Their profits get reinvested into the network instead of pocketed by some CEO who doesn't do any work.

Private, for-profit companies are inherently designed to offer the lowest possible quality for the highest possible price, which ultimately means the consumer gets fucked - especially if there is no real competition.

I did not specify the government. Nor did my statements rule them as the only option left. Could be a non-profit organization of some sort. With that said, as someone else pointed out, municipal ISPs do have way better customer satisfaction.

And ISPs should be defined as a utility. And they shouldn't be allowed to be private, for-profit companies. But we are where we are.

On one hand, I should kinda care because payment processors should not have that much power. On the other hand, the only thing we're losing is low-quality, low-effort, fetish porn "games" - many of which probably barely count as a game at all. So, I'll consider it a wash for now.

They're already taking the picture. It wouldn't be hard to carry a patch or something like that they can put in the picture before taking it. Takes two seconds and kills two birds with one stone.

Advertising that the police are inept and using technology that they should never be using. Ever.