Proposal for CPUs/graphic cards for PCs power draw standardisation

[OneMoar]

No lol just No so much No no no no PLEASE GO NO no Nooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

 

[BoggledBeagle]

No lol just No so much No no no no PLEASE GO NO no Noooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo


The difference between stupidity and genius is that genius has its limits.
~Albert Einstein

Limits, you say, or cite. Oh yes, limits. Continuum of dark stupidity without borders and small limited islands of ingenuity, I guess your post was to illustrate the continuum?


Think and imagine:

One can optimise and build a quality product only in a well defined and limited space.

Optimisation in general is possible only if constraints are present.

First you need an anchor and then you can build a boat.

Thick fog of devillish sulphur vapours of marketing lies and misleading specs ruthlessly shone through by defined intense light beams of truth and reason, blinding and eventualetty killing those devils who produced such inhospitable atmosphere intended to deceit.

THINK AND MAKE ALBERT HAPPY!

 

[Wirko]

Defining clear product categories with one fixed paramater greatly facilitates choosing products for consumers, because they can easilly judge performance of these products. Everything about performance of the products is more transparent and clear.

It wouldn't even work for dishwashers. Which have just one use case, with a few small variations.

 

[BoggledBeagle]

It wouldn't even work for dishwashers. Which have just one use case, with a few small variations.

Home dishwashers are all the same size with very similar performance, and making different classes of them is pointless. One CPU on the consumer PC market can be six times larger than the other one, measured in core count.

Dishwashers and all other home electrical appliences have a true power draw specification, because you need to know what electrical breaker to use and how the appliance loads your whole electric installation in your home and what appliances you can run concurrently. In case of AC units you have also a starting current spec, so that you have all the information about the true requirements of the appliance to be able to run it safely and reliably.

CPUs not only consume different amount of energy (one CPU on the consumer PC market can draw 200W or even more than the other one) and you need to take that in consideration while choosing a power supply, you also need to cool them, coolers are different sizes and some are simply incapadle to work with certain CPUs. Knowing a true power draw of a CPUs is therefore twice as important than knowing a true power draw of a home appliance.

Efficiency of home appliances is rated and extensively tested: e.g. you test energy consumption and water consumption per washing cycle of some standardised clothes load.

There is really no reason why any computing components should be treated differently than any electric home appliances, why consumers should have less information, why should they be misled and enjoy less consumer protection.

Your PC while gaming can easily draw say 500W, you can run 10 fridges at the same time with this sort of power !!! TEN FREAKING FRIDGES WITH THOSE COLOURFUL EFFICIENCY LABELS ON THEM, WHICH MANUFACTURES HAVE BEEN TRYING HARD TO MAKE AS GOOD AND EFFICIENT AS POSSIBLE. You can run a heatpump / AC unit for a whole home with the power some PCs consume.

Also, electric home appliances tend to be optimised and balanced products. Can you imagine how a washing machine would behave it is was set out the box as Intel 14900K, for example? BTW I am not an Intel hater, and have this CPU and I am happy with it, once I adjusted it.

 

[BoggledBeagle]

I just watched this video about the new Nvidia GPU power connector:

It seems that the safety margin of the new connector is lower than that of the old 8 pin design, and the connector has a real limit of 660W, while a specified maximum is 600W.

I just want to note that this new connector is perfectly OK and adequate in the case consumer GPUs were limited to power draw of 250W, beside all the advantages that defined classes of GPUs based on real power draw would bring.

One would have 4 categories of GPUs (65, 100, 160 and 250W) to choose from, and in each of these categories you would be choosing between highly optimised and energy efficient cards, which would be easily comparable between each other, intensifying competition and thus bringing these excellent products to customers for the lowest prices possible.

Please compare what I just wrote to the current mess on the GPU market. Do you believe that the manufacturers are really competing between each other and selling their products at the lowest possible prices? Do you really believe that each product you see on the market is highly optimised and efficient?

 

[BoggledBeagle]

It could be useful to think once more about what I proposed here. There is a 600W GPU ready to hit the consumer market.

 

[_JP_]

Right, but what you're looking for is something akin to the EU Energy labels, which would only make sense by a long-shot through a thin straw for OEM pre-built devices. And even then, maybe laptops with no upgradability would be the only eligible class, but those are e-waste anyway.
Because unlike an appliance, multimedia unit, lamp, a tire, etc, computers aren't closed systems that once placed for regulatory approval, have clear measuring guidelines to determine their maximum, minimum and stand-by (if existing) consumptions and then calculate the consumption on average per year, which is what is considered to place a "score" on the efficiency.
And unlike a house (like in Portugal), it doesn't understandingly stay with the same configuration for 10 years for it to have an regulatory energy evaluation to consider its energy efficiency as a structure/system.

Computers are as efficient as their purpose (read: software ran on), which varies widely, not to mention all the part that are interchangeable that can wildly modify said efficiency, not to mention how fast that can happen.
It's not by chance that since 2016, forums are filled with sweat-hours of discussions regarding 1% lows in FPS performance per game, as an example.

Like others have mentioned in the past, it's up to you to read the technical data if that is what you're looking to figure-out and measure what components you actually need to build a computer. There are standards, but not in the marketing/PR vomit for which the only purpose is to erode your inhibition to make a purchase (and it's really well done, it's an impulse buy).

 

[BoggledBeagle]

Right, but what you're looking for is something akin to the EU Energy labels, which would only make sense by a long-shot through a thin straw for OEM pre-built devices. And even then, maybe laptops with no upgradability would be the only eligible class, but those are e-waste anyway.

No, that is not what I am looking for. Try to read this again and understand it.

One would have 4 categories of GPUs (65, 100, 160 and 250W) to choose from, and in each of these categories you would be choosing between highly optimised and energy efficient cards, which would be easily comparable between each other, intensifying competition and thus bringing these excellent products to customers for the lowest prices possible.

Please compare what I just wrote to the current mess on the GPU market. Do you believe that the manufacturers are really competing between each other and selling their products at the lowest possible prices? Do you really believe that each product you see on the market is highly optimised and efficient?

Competition is not a given thing. Sometimes companies intentionally avoid it in order to be able to charge higher prices for their products.

If there were two categories of clearly defined enthusiast GPUs, competition would be unavoidable. You have a clearly defined goal, for example: bring as much performance at the constraint of 250W. You cannot make a card, which consumes more, because it is banned. You cannot make a card, which consumes something between the limits, because it would disadvantage you in comparison with the card that at this limit. You are pinned to the wall, you do not have any wiggle room, you cannot lie, you cannot misrepresent anything, you simply must do the best you can to win a sale against a competitor in a given category.

Runners can be compared, only when there are clearly defined race lenghts.

Competition is necessary for consumers to get good product for good prices. When competition lacks, prices need to be regulated. For exemple if AMD and Intel would be unable to compete in a 250W GPU category in some time period, the only remaining producer Nvidia should have the prices of their products in this category strictly regulated.

 

[_JP_]

No, I did read it but went for the bigger picture, because what you're specifically requesting for just results in stagnation (see: FIA motorports competitions).

The difference with that example is that in motorsports, eventually the regulations change after 1+N years, because it has stagnated and viewership dropped or "ooops", somebody actually found a way to have an advantage. The regulation will move to ban that advantage and keep the stagnation the competition, although the latter situation is my definition of competition. It's indirectly linked to the consumption part, the viewership.

So what you're asking for requires a regulating body, which already happens for Telecom/ISPs. The regulating body defines rules and used standards, arbitrary authorizations, but never how companies are actually to compete, otherwise it would not be a free market.
If you're familiar with that realm, you'll know how imbalanced or cartelized it is between ISPs (you can look at my country, as an example). And these are directly linked to the consumption part, subscribers.

So the consumer would actually not benefit from that, at all.
And again, at the pace technology moves, regulating bodies are already considered dead weight regarding the time they take to create the materials to place rules on top of standards. (even more constrasting with my previous example, except technology does stick around for more than a championship season, unless it's a market faliure from the get go.)

Also, who says I'm okay with just 250W for a GPU if 600W is what enables X performance with the current technology?

There is already a hidden motivator for the competitor that comes after with a product that does X performance but at 450W (for example), since the consumer might as well go with the lowest running costs per year, if the initial investment is the same.

You might argue: "But NAND/RAM makers were caught not competing!"
Sure, they were. But the market didn't give a damn and the fines were minimal, so only economic downturn hurts and in turn could break that cartel.

Still, nothing beats you learning more about the matter, actually doing your homework, using your wallet as a competition enabler, even if that means sitting put or not going in the impulse of the latest greatest tech.

 

[Bill_Bright]

It could be useful to think once more about what I proposed here.

Useful to whom? Consumers? IMO, what you propose will stifle competition, suppress free enterprise, and the people's freedom of choice!

Who is going to impose those limits, the government? Even those of us who think government regulations are necessary evils would likely be opposed to that.

Are you going to fine AMD if their system demands 162W? Who is going to inspect and certify these products? Are you going to impose increased taxes on the taxpayer to fund these investigatory agencies so they can put watchdogs in these businesses? Are you going to pay for the necessary resources to do their jobs - not to mention pay law enforcement to enforce those regulations when violated?

Who is going to pay for all that? The manufacturers? Yeah right. They will pass those costs along to us. Taxpayers? Again that's us footing the bill.

I get your point but it is not realistic. Why? Because this is not about how much weight per square inch a certain formula of concrete can support in a multilevel parking garage. This is not about how much current a certain gauge wire can carry in a high voltage circuit. This is not about the impact resistance of a pane of glass in the hurricane prone coast of Florida.

You are not trying to impose necessary safety standards. I'd be all for that. You are trying to unnecessarily limit the performance of my computer! How dare you? (Rhetorically speaking)!

Can you imagine the public uproar if you imposed such limits on the automobile industry limiting engine size and horsepower? Look how hard it has been just to get the industry (AND the public) to embrace better fuel economy! And even then they are kicking, scratching and clawing every inch of the way.

Competition is necessary for consumers to get good product for good prices.

Exactly. But you want to dictate standards by putting limits on the number of options, and the maximum level for each of those options.

What "fun" would there be for enthusiasts if YOU destroyed car or boat racing because you put a cap on horsepower and top speeds? What incentive would the automobile industry have to be innovative, to advance technologies, to be competitive if you restricted their options to such a limited handful of options?

IMO, you dictating such specific categories/levels would stifle innovation and thus competition! It would stifle the fun and excite and competition among the manufacturers and enthusiasts who, for the pure enjoyment of it (plus bragging rights), work hard to eke out 2 more FPS or 10 more clock cycles.

What incentive will board makers have to create different options for us consumers if you put a limit of 5 options on them (25, 40, 65, 100 and 160W)?

I say let the market do the regulating - that is, let us consumers dictate what the manufacturers produce with our purchasing power - not through government regulations.

 

[lilhasselhoffer]

Limits, you say, or cite. Oh yes, limits. Continuum of dark stupidity without borders and small limited islands of ingenuity, I guess your post was to illustrate the continuum?


Think and imagine:

One can optimise and build a quality product only in a well defined and limited space.

Optimisation in general is possible only if constraints are present.

First you need an anchor and then you can build a boat.

Thick fog of devillish sulphur vapours of marketing lies and misleading specs ruthlessly shone through by defined intense light beams of truth and reason, blinding and eventualetty killing those devils who produced such inhospitable atmosphere intended to deceit.

THINK AND MAKE ALBERT HAPPY!

So....I read this, and now understand that you have no idea which end is up.

As an example:
1) In my experience most sail boats have no anchor.
2) Ships can have more than one anchor.
3) Anchor size does not mean a thing when you have no idea what the constituency of the bottom is.


Translating this into similar things for computers, only quickly;
1) The processor can not have any cooler at all. It can also install a fan, which means it can have a smaller one. Physics is a rough concept to wrap your brain around, but to artificially impose the need for a thing is to constrain what a thing can be. If you want a car example, most have flywheels...except the F1. It's basically the first hypercar, and it completely removed the flywheel.
2) This is harder to conceptualize with a CPU...so the best I can do is that liquid coolers transfer heat from one location to another, and thus can be huge without a huge footprint on the CPU.
3) A CPU rated by thermally dissipated power does not know ambient temperature. It has thermal fusing and limits. You can theoretically run a 65 watt CPU without a fan in a room with an ambient temperature of freezing, and you can also run it at an ambient of 60 degrees Celsius. This is why wattage is a measurement of energy transfer, not a measure of change in temperature, but there is a hard limit to physical failure.


If you wanted to you could forego all of the ability to boost without overclocking, define a single ambient temperature and room volume, define a CPU loading that magically was a certain known percent efficient, and then do all of this to "optimize" a CPU. It's basically the definition of chasing nothing, because by defining such a narrow condition you have made it impossible to exist in reality, and thus made all of that work useless. Nobody can stop you from doing that...but nobody has to agree to such a useless effort either.

So you can understand, your premise is the same as a sore loser child on the playground. I may not be the fastest, the strongest, or the smartest. Despite all that, I'm still the best because I can run 300 feet backwards while holding a half cup of water steady enough not to spill any the fastest. OK little homie, just let it go. We know you don't understand that this isn't a useful metric, but it is yours because you do it best...until somebody else comes along and does it better, then it'll also include balancing an umbrella, and running on the 4th Sunday of February as a requirement.

 

[DirtyDingusMcgee]

Who are you proposing this to exactly? Is IEEE, UL, NEMA or ECES going to see this and think "this guy is on to something, let's push radical changes to how all these huge companies design their products"?

Why tell us? None of us are in a position to do anything about it. It's kind of just mental masturbation.

And what if we, and the consumer base at large, like it the way things are? Why should this be imposed upon us or the manufacturers?

 

[_JP_]

Still picking-up on the example of the NAND/RAM manufacturers not competing.
If stricly regulation is the way you feel comfortable with, then something to the tune of a regulatory body of markets and trades that would go after corporate aquisitions, mergers and post-merger termination of brands/entities is something that could affect the constant non-competiveness you feel that exists.
To put it quite simply, like you mentioned, for GPUs you have AMD, Intel and nVidia, but there used to be more.

That regulatory body needs to be backed-up by laws that stop companies from anulling their competitors and have enough reach to curtail the current too open definition of "enough competition, thus not a monopoly" rulings that allow acquisitions to still take place.

Now, countries do have that regulatory body, I know. I was just placing it vague enough as a writing style, but in most cases, those regulators still allow aglutinations to happen.

In the NAND/RAM makers' example, the issue mostly stems from not existing enough players anymore in the market to compete.

The exception I see is something like Stellantis, in the automobile world, that bought so many brands/manufaturers that couldn't terminate or transform them, so ended-up competing against itself and now we're just waiting for it to go supernova.

Not saying conglomerates shouldn't exist, or parent companies, so on, but the width and connections AND reach could be better looked at.

Who are you proposing this to exactly? Is IEEE, UL, NEMA or ECES going to see this and think "this guy is on to something, let's push radical changes to how all these huge companies design their products"?

Why tell us? None of us are in a position to do anything about it. It's kind of just mental masturbation.

And what if we, and the consumer base at large like it the way things are? Why should this be imposed upon us or the manufacturers?

I would not put aside that on this forum, someone actually is a part of or will be a part of a body, that could see these threads, and then we actually exert some influence, by way of opinion. This is a forum, after all.

 

[TheinsanegamerN]

THINK AND MAKE ALBERT HAPPY!

Here's a "think" for you: every single time restrictions are put in place, they are always tightened, to the detriment of the consumers.

Why would we allow 160w, or 100w? Or even 45w? why not restrict all consumers to 25w CPUs and 75w GPUs? Actually, why even allow GPUs? Gaming is a wasteful hobby that produces nothing of value, you shouldnt be buying dGPUs for that! Think of the CO2 savings! you dont need more than an AMD G series to play bing bing wahoo!

See how easy that is? You REALLY think industries and political groups wouldn't do that?

Also, what you consider sufficient performance may not be sufficient for others. If I can justify and pay for the energy to run higher TDPs, that is my choice as a consumer. Why would I voluntarily agree to a set of standards that limit what I, the consumer, can purchase? Why would the industry agree to this?

 

[DirtyDingusMcgee]

I would not put aside that on this forum, someone actually is a part of or will be a part of a body, that could see these threads, and then we actually exert some influence, by way of opinion. This is a forum, after all.

I was trying to find a nice way to say "this crazy shit won't happen"

 

[Bill_Bright]

Here's a "think" for you: every single time restrictions are put in place, they are always tightened, to the detriment of the consumers.

Except that is not true. Some times it is, but not always.

On the flip side, many times when restrictions are eased or removed, it typically is to the detriment of consumers too - big business (like banks, insurance companies) typically win.

 

[TheinsanegamerN]

Except that is not true. Some times it is, but not always.

On the flip side, many times when restrictions are eased or removed, it typically is to the detriment of consumers too - big business (like banks, insurance companies) typically win.

Those restrictions far too often backfire, removing options for the consumer.

Small fuel efficient trucks, small cheap eco cars, appliances that would run for 50+ years, all eliminated from our market in the name of "efficiency". Some models of Dell systems are not allowed in California due to efficiency regulations. It's a miracle they havent banned motorcycles yet.

I would much rather deal with a market where consumer have choice, but have to undervolt their cards for the most efficient performance, over a market where you are limited to nothing more then RX 6400s because the government said so.

 

[Bobaganoosh]

You cannot make a card, which consumes more, because it is banned.

Live free or die.

 

[ShrimpBrime]

I'm happy reaching big wattage.

Here's 400w CPU ok CBR23.

Restrict deez nuts.

 

[Bill_Bright]

It's a miracle they havent banned motorcycles yet.

I don't see that happening any time soon. Even gas-guzzling motorcycles are typically more fuel efficient than most regular cars - including many hybrids.

Small engines (lawn mowers, leaf blowers, small generators, etc.) are another story - though it is not about efficiency but the massive amounts of pollution they spew about.