_nist

As a frequent work traveler, airport prices are always at least double what they are outside the airport. It's not a fair comparison.

It depends on what you mean by "more likely to announce things with no industrial relevance". If we are including humanities, then yes. If we exclude humanities, then no. Nearly all current STEM advancements can be traced back to early university work because industries normally do not pay for R&D because of its high risk, low reward effort. Not to mention, it's a money sink. Even current AI models come straight from university work, intermixed with private industry who often apply these concepts and scale them, but aren't the idea originators.

Companies lie and exaggerate all the time. It's called PR. It makes no difference whether it comes from a company or university, but companies have much more incentive to lie than universities, especially in technical fields, because they don't have to subject their claims to peer review unlike universities.

Doesn't surprise me in the least. McDonald's is such a bad company on many levels, even from its founding roots. They make fast food (an industry notorious for craping all over their employees) and it was usurped by a greedy business man (Ray Kroc) from its original founders. Their food is okay, but I would never, ever do work for them.

I was addicted to these games as a kid. Too bad they never remade them and now its some obscure title nobody can get.

It would be anti-trust if they were deliberately shrinking supply to hike up prices on their cards. For example, I know I have the best card in the industry, so I'm going to force the supply low so I can charge effectively whatever I want.

It wasn't in the 2000's back when ATI and NVIDIA cards had comparable prices. Sometimes ATI cost a little more because they were better, but they were never crazy different from NVIDIA. And there were never supply problems.

Microsoft delivers AI recall Everyone hates it Apple integrates AI into every facet of a device that is highly personal Everyone loves it

Please make it make sense.

Yeah, it's a shame, I often have to do this type of crap in my off time. However, having a well architected app significantly reduces these types of massive undertakings. Doing it right the first time has its advantages. Then again, weighing it against delivering early and other important aspects has its merit too. It's just a double edged sword unfortunately that you often have to walk in this industry.

Output has been scarce for NVIDIA gpu's for quite some time. Often not even being able to support the demand for their cards. NVIDIA is a trillion dollar company right now. There should be no reason why they are restricting access to their reference cards and not able to support demand.

You are also not counting handheld use such as the Steamdeck. There is a reason Steamdeck doesn't use NVIDIA graphics.

When AMD can run 95% of games as good as any NVIDIA card and you factor in that most games have virtually no ray tracing support or any other fancy feature that NVIDIA offers, coupled with the known low performance/cost gains in their own graphics family, and their lack of ability to support demand despite being a trillion dollar company, I'd say that there isn't much of a leg to stand on.

I understand antitrust law enough to know that "mad I can't game on Linux" is an anti-competitive measure.

As far as what I read about the driver stack, they only recently (within a year) release source for kernel modules. They did not open source the whole stack. And these are only new modules in an alpha state.

Its basically the bare minimum.

Most people aren't running ML on their workstations and these mostly use entirely different types of cards.

This is one successful landing. I think its robustness needs a lot more successful landings than this. Plus, its not even landing like it was originally designed to land, its just landing in the ocean like any other rocket expect with thrust vector control rather than parachutes.

I'm not sure what world you live in, but Nvidia sells their cards at 1.5 - 2 times the price than AMD's on average and AMD has had strong Linux support for years.

I agree with this. In a large organization, if you have risen to a level where you are being relied upon at a regular intervals, it is imperative that you have a well architected solution that you can readily change and this is where you separate your program from spaghetti code to something useful. Sure, its nice to write unmaintainable junk when toying around, but I to have seen too many codebases where people were just throwing features in without thought and it causes the program to become way too constrained to only a specific problem domain and it becomes inflexible for solving new problems (to the point you have to re-write nearly everything from scratch).