NVIDIA May Be Trying To Prevent GeForce GPUs From Being Used In Data Centers

[coder]

The problem we have today is that too many people are investing in companies with the idea that "making as much money as possible" is the goal

I would extend this, applying it to capitalism, in general. When the public appreciates the how's and why's of the proper role of capitalism in society, they're more likely to support laws and regulations ensuring that corporations are subservient to the interests of the public rather than vice versa.

and when those people end up in a majority ownership position they distort the guidance given to management, which in turn leads to short-horizon decision making, out-of-control off-shoring and "right-sizing", debasing of respected brand names, unethical/illegal behaviour and a host other things we would be better off without.

We need to draw a line between "reasonable reward" and "greed" - greed is what makes capitalism fail over time, not what makes it succeed.

It's a nice idea, but too fuzzy to implement. I would focus on specifics, like accounting rules, reporting requirements, tax incentives, varying capital gains tax rates for different types of investments, and dis-incentivizing short-term investments.

The root of the problem is that the public understanding of capitalism, corporations, investment, and the economy is too simplistic and lacking in essential nuances. So, there's little support for lawmakers trying to do the right thing for the people, and too much pushing for them to be overly generous (mostly) or punitive (rarely). I'm not optimistic, but it's hard to fix a problem without first understanding it. If we can at least do that much, perhaps there's reason to hope.

[bridgman]

Corporations, on the other hand, are motivated solely and pathologically by greed. It's in their charter. It's their only mandate. If shareholders see management failing to maximize the profits and/or the well-being of a corporation, CEOs get replaced and lawsuits get filed.

No, there is nothing in either typical corporate charters or in corporate law which aligns a corporation's purpose with greed. It is not their only mandate except to the extent that shareholders dictate it. For-profit corporations are *allowed* to generate a profit (as opposed to non-profit corporations) and operate under different laws as a consequence.

A corporation is nothing more than a group of people authorized to act as a single entity.

Not that I'm opposed to corporations - that focus can be vital to their effectiveness. But we should see that they act in ways that would rightly earn any human the label of an extreme sociopath. So, it's necessary to bound their behavior with regulatory guard rails to harness their greed in service of society and be wary of their attempts to co-opt society in service of their interests.

No argument here - as our societal ethics degrade (I was going to say "social and political ethics" but it's possible that political ethics have always been at rock bottom) it becomes harder to rely on people considering the greater good, and we end up having to fall back on laws. Unfortunately the same people who cause the problems in the first place will also view each new law as something to be worked around rather than as a guideline to expected behaviour, so we end up with an ongoing battle between increasingly rich individuals and increasingly poor governments.

The root of the problem is that the public understanding of capitalism, corporations, investment, and the economy is too simplistic and lacking in essential nuances. So, there's little support for lawmakers trying to do the right thing for the people, and too much pushing for them to be overly generous (mostly) or punitive (rarely). I'm not optimistic, but it's hard to fix a problem without first understanding it. If we can at least do that much, perhaps there's reason to hope.

The other challenge is that as you say public understanding of the economy has not kept up with its increasing complexity, and as a result the electorate has effectively become more gullible over the years, so anyone trying to promote a better economic model is going to have a very tough time selling into any part of the electorate after 30 years of politicians claiming that the key to good jobs and a stable economy was either greed-driven capitalism and unrestricted free trade or unrestricted growth in government. Neither model really works, and so western economies are increasingly supported by government deficits, consumer debt and artificially low interest rates.

I am starting to believe that the presence of royalty may have helped to keep politicians a bit more honest in the past (since royalty is by definition forced to take the long term view) but it's not clear what the modern equivalent of that would be.

It may be that the natural cycle is the one that everyone keeps fighting - "true" capitalism (look at founder-run companies for examples) -> greed-driven capitalism -> collapse into communism -> emergence of dictatorship -> technocracy -> next chance at "true" capitalism.

The problem is that it's all too easy for even that cycle to skip over true capitalism and jump straight to "greed-driven" again - it's totally a function of how skilled the leaders are at the technocracy stage, and whether they are willing to impose sufficiently high penalties (eg death) in order to send a clear message about how leaders are expected to behave.

The closest thing I have seen to a working, repeatable model is the one that South Korea was using when I was working there in the 80's. I think it had been more or less in place since the 60s though. The model as I understood it was that the big family-run conglomerates (the chaebols) were free to be as successful and profitable as they could be, subject to a few hard rules:

- if someone wanted to work, the companies had to find jobs for them (even if that meant a lot of elevator attendants initially) - and over time companies learned to make the most effective use of people they were required to hire anyways

- when companies set up big new facilities they also had to pay for the associated infrastructure including housing for employees

- some products (switching power supplies were the example I encountered) could only be manufactured by companies under a certain size (75 employees maybe ?) and the larger companies had to deal with them rather than going outside the country

It wasn't perfect (in some ways it was only workable because SK had a small number of massive corporations at the start) but it was still the closest thing I have seen to a workable model. Doesn't have to be that exact set of rules, but I do think it will take more than just tax laws (or taxation in general) to get things back on track.

[boboviz]

I'm a boinc boy. They want to sell their gpgpu card. But a lot of scientific projects (like gpugrid, einstein@home, seti, etc) use the volunteers geforce.
Go to Amd and Opencl!

[aht0]

1. budgeting isn't a choice taken by some god-like figure that decides where to spend money for all departments down to white A4 paper and printer supplies, and is also all-knowing enough to make wise choices on everything. The budget is split to each department's head manager, and then depending on how it goes it may be split more (may have sub-departments or something else). Those that decide to buy hardware for their department or subdepartment only have a relatively small amount of money, which usually covers the expenses and salaries and little more.

I doubt that our hypothetical researcher is unique in her/his needs for computational power.
Thus it seems idiotic fragmentation of resources allocating funds for setting up dedicated computing resources for a single project. Instead of having central much bigger data center for a whole university from where researcher can draw from at need or as per allocated time.

2. GPUs are a tool, a means to an end. When you gotta get a job done, you don't care of stickers. On quite a bit of computing loads you don't need ECC VRAM or higher buffers or whatever, you only need raw power, low power consumption and other more mundane things. And if that is the point, then it is simply stupid to buy more expensive cards just because you are big and wealthy (and stupid).

Professional cards have way better double precision FP performance, which is important in engineering and scientific calculations. At least tends to be more important than single precision FP performance consumer cards are good for. Why is ECC unimportant in scientific calculations? I'd assume possible errors are something you definitely do not want.

[aht0]

Unless they actually defined "Data Center", then you should be aware that creative lawyers can and will bend that definition to include things like server rooms or even potentially computer labs. You need to dig into the meat of the specific language they used before deciding whether it potentially applies to specific universities and departments.

It might also simply be as generic as possible because local law may rule out bunch of meanings and nVidia's lawyers tried to get the max out of the wording.

[duby229]

- some products (switching power supplies were the example I encountered) could only be manufactured by companies under a certain size (75 employees maybe ?) and the larger companies had to deal with them rather than going outside the country

It wasn't perfect (in some ways it was only workable because SK had a small number of massive corporations at the start) but it was still the closest thing I have seen to a workable model. Doesn't have to be that exact set of rules, but I do think it will take more than just tax laws (or taxation in general) to get things back on track.

That sounds like protectionism to me. The US already knows all about that scenario. IMO it was the US's protectionist policies that lead directly to the dustbowl. And indirectly lead to monopolistic corporations that helped crash the economy.I know the direct cause of the great depression was a lack of government regulation, but that indirectly lead to powerfull monopolies. If SK didn't have strong allied countries it had to trade with there would be no possible way it could work. Strong monopolies always develop eventually into terrible caste systems.

[bridgman]

Definitely an element of protectionism, at least to the extent of considering each countries size and wage/price structure when establishing trade rules, but IMO some of that is going to be required anyways.

SK seemed to have pretty reasonable trade policies when I was there - tariff barriers on specific products were steep but time-limited... basically "you have this long to get your industry established then you're on your own".

The reason I find SK an interesting model is that they were able to make drastic strategic changes in their economy without requiring a dictatorship to make it happen. Instead they were able to align corporate interests with societal interests and then let everything run within that framework.

In North America we fail miserably at that. Europe is somewhere in between.

[coder]

The reason I find SK an interesting model is that they were able to make drastic strategic changes in their economy without requiring a dictatorship to make it happen. Instead they were able to align corporate interests with societal interests and then let everything run within that framework.

It was something of a dictatorship until the late 1980's, IIRC. It's also ethnically and culturally very homogeneous, which is associated with higher levels of societal trust.

[Spooktra]

Making the most money possible is called "greed", and the whole capitalistic system is based on what is basically "greed".

You could have just saved yourself a bunch of typing and just admitted that you are a commie at heart, the views you have expressed are typical leftist, Marxist inspired socialist, Bernie loving commie crap.

I have a few minutes to waste so I'm going to try and reason with you.

You want a job where you get steady raises, health benefits, good O/T, good PTO and sick days, bonuses and steady employment / job security? Well sorry to rain on your commie parade but the only way a company can offer that is by making as much money as possible. A business who's primary goal is not to make as much money as possible is called a charity.

Would you rather work for a company like AMD or Intel? Would you rather work for a company that can afford to hire and keep new employees or would you rather work for a company that keeps announcing job cuts and other cost saving measures?

If you ever open up a business feel free to embrace your hippie, leftist, commie business ideology and drive your business in bankruptcy and lay off your employees, I'll just make as much money as I possibly can and make sure my employees have steady, good paying jobs, and that my business's bank account is as fat as possible.

[Spooktra]

Please stop parroting this before another generation grows up with a corrupted "me me me" view of the world and makes things even worse.

The purpose of a business is to provide goods and services on a larger scale than is possible for an individual, and while doing this requires a certain amount of gross profit (for salaries, bonuses, benefits, R&D etc..) and net profit (interest on borrowed funds, return on invested funds, supporting growth) the primary purpose of a business is still the net value it adds to society.

The problem we have today is that too many people are investing in companies with the idea that "making as much money as possible" is the goal, and when those people end up in a majority ownership position they distort the guidance given to management, which in turn leads to short-horizon decision making, out-of-control off-shoring and "right-sizing", debasing of respected brand names, unethical/illegal behaviour and a host other things we would be better off without.

We need to draw a line between "reasonable reward" and "greed" - greed is what makes capitalism fail over time, not what makes it succeed.

I want to thank you for this, I always thought that the reason AMD was an also ran, a company that was destined to be in second place perpetually had something to do with AMD's acquisition of ATI but you made me realize that I was wrong all along.

I want to make this clear that I am not trolling with what I am about to say, I am as serious as can be, you work for AMD correct? If I was your boss and I saw that you said the following:

"the primary purpose of a business is still the net value it adds to society"
"We need to draw a line between "reasonable reward" and "greed" - greed is what makes capitalism fail over time, not what makes it succeed"

I would fire you immediately, you would no longer be an AMD employee and I mean that sincerely.

Talk about a distorted view of reality!

You think that capitalism fails over time? Name on time when it's failed? The capitalist system is based on "let the market decide" and "survival of the fittest", when businesses fail and investors lose their money it's because the capitalist system work, not failed.

Further, you think that "the primary purpose of a business is still the net value it adds to society"? What is the net value you see AMD adding to society? What is the net benefit Burger King, McDonald's, any fast food restaurant, any fine dining restaurant, any gas station, any bar, any liquor store, any small mom and pop outfit, what is the net benefit you think they add, other than providing steady work to its employees?

I am going to bookmark your post and anytime someone wonders why Intel is worth billions and AMD just barely had a company wide quarter in the black, I will link to your post and everything will be explained.

I feel sorry for AMD, if they employ people with your economic view points I predict they will be filing for bankruptcy protection within 2 years and will be taken over in an asset acquisition within 3 years.

I really hope people stop buying AMD products, clearly the corporate culture over there is tainted with commie ideology and needs to be allowed to go the way of 3dfx.