Intel ATX 3.0 16-pin Power Connector for PCIe Gen5 is Smart, Has Four Power-Delivery Variants

[Valantar]

what i tried to show you is the pertinent information to relievated your concern over the sense wires.

I know, and I appreciate the input, but the issue is that what you're trying to address is not really pertinent to the issue I'm bringing up. You're talking standards and sensible applications, I'm talking real-world, nitty gritty realities of how these standards are botched by unscrupulous people out to make a quick buck.

well you said it yourself, cheap. and sure i myself have seen forum horror stories of extensions, even seen seasonic pci-e connectors melt. doesn't mean it common.

No, but is that an argument for not criticizing a standard that incentivizes poor implementations? I would say no.

don't confuse laissez-faire with confidence from knowledge and experience buddy.

Confidence from knowledge of and experience with what? Seemingly not the cheap custom PSU wiring market, given the approach here - your working assumption seems to be that there are good actors in the market, so the bad ones can just be ignored. I'm very much opposed to that line of thinking. And in order to minimize the potential for harm from these bad actors, what we need is for standards to remove opportunities for potentially dangerous shortcuts, not add more.

oh my, blatant security issue??? yeah no, you are so waaay blowing this out of proportion. and i would point out that those safety features are there in part because of the prevalence of crappy extensions, adapters etc that are used for broad range of implications even though intel and psi-e sig does specify exactly how things are implemented ya know, standards. and its like you don't think anyone from intel, or pci-e sig ever shopped on newegg, amazon, ebay. . . .

Blatant security issue might be a bit strongly worded, but IMO it is a major design flaw when your design actively incentivizes shortcuts and disincentivizes compliant implementations, even in niche applications. That is a severe regression in quality for this standard vs. previous ones.

How? It was brought up earlier that similar shortcuts could be done with 8-pin PCIe wiring, as those also just need sense pins shorted to ground. But I have never seen this done by anyone, ever in real life. Why? This is speculation, of course, but it's relatively safe to assume that when you're already making a harness of 6 wires, adding two more identical wires is a low-effort undertaking, and coming up with and implementing a shortcut would likely take far more effort than it would be worth. It might even look worse in the end, as 8 matched wires likely look better than 6 matched wires + a couple of bodge wires in the plugs. And, as has been central to my argument all along: these people care mainly about aesthetics.

For these cables, on the other hand, there is no real option for matched cabling (unless, as I tried bringing up as a possibility earlier that you apparently misread, you could find pins for the small connector capable of holding wires of the same gauge as the power wiring - this would let you run 16 equally thick wires for aesthetic purposes, at the cost of unnecessary wiring, exotic and possibly nonexistent pins, and a lot of hassle). You'd then need to find a way to make having four much thinner wires look good alongside 12 thicker ones. How do you do that? Creative routing and combs might help, but that is labor intensive and increases QC costs, which is anathema to these actors. Two-conductor wire of rougly the same thickness as the thicker wire might work, but terminating them in a good-looking way would be a challenge. Or you could stuff two wires into one sleeve - though that would be lumpy and stiff, sub-optimal for aesthetics. The same goes for sticking single wires into oversized sleeves - plus that they would be loose, of course. The third, and suddenly more attractive option, then becomes shorting the wires to ground at the connector. This would be the simplest, cheapest (you're now saving four lengths of wire per cable, rather than the potential two on an 8-pin), and possibly least bodged-together-looking solution of the bunch. That is a problem, when the worst implementation possible has the most advantages.

for the love of all that is holy, have you not understood a thing??? - they are not full size terminals

............................................. I was comparing the pros and cons of going with 16 equally sized pins vs. the solution they actually went for. Was that difficult to understand? Literally the entire basis of my issue here is based on these being different, smaller pins, ffs. I would really suggest you take a step back here, as it seems you're approaching this discussion dead-set on just not understanding what I'm saying, whether consciously or not. I'm sorry if my style of writing isn't sufficiently clear, but literally nothing in the sentence you quoted above is me saying that the pins on the standard presented here are full size.

i've never said anything about what people deserve and there is no place for that in a reply to me. however, i will question the common sense in buying a $5 cable for a $3000+ card.

And I'd agree with that assessment. But I'm also realistic enough to know that people will still do so, and that people fall victim to scammers passing off $5 cables for $100 cables, but with an impressive rebate, etc. That's why standards should prioritize safety to the degree that eliminating unscrupulous shortcut implementation possibilities should be a main goal of the design. That's how you make a safe standard - by not assuming that everyone implementing it is a good actor interested in following it. Instead, what we have here is a standard that, though unintentionally, at least partly incentivizes shortcuts.

 

[chrcoluk]

Another thing I noticed is in that spec sheet, nearly all of the graded PSU sizes have 300w allocated to CPU's, only the bottom one is reduced to 275w, so what does that say about future intel CPUs?

It compromises GPU capacity, so on 3.0 a 750w power supply can only handle 300w for GPU instead of more like it can do now.

Also UPS capacity now might be an issue as well with the 200% burst requirement.

 

[looniam]

I know, and I appreciate the input, but the issue is that what you're trying to address is not really pertinent to the issue I'm bringing up. You're talking standards and sensible applications, I'm talking real-world, nitty gritty realities of how these standards are botched by unscrupulous people out to make a quick buck.

No, but is that an argument for not criticizing a standard that incentivizes poor implementations? I would say no.

Confidence from knowledge of and experience with what? Seemingly not the cheap custom PSU wiring market, given the approach here - your working assumption seems to be that there are good actors in the market, so the bad ones can just be ignored. I'm very much opposed to that line of thinking. And in order to minimize the potential for harm from these bad actors, what we need is for standards to remove opportunities for potentially dangerous shortcuts, not add more.

Blatant security issue might be a bit strongly worded, but IMO it is a major design flaw when your design actively incentivizes shortcuts and disincentivizes compliant implementations, even in niche applications. That is a severe regression in quality for this standard vs. previous ones.

How? It was brought up earlier that similar shortcuts could be done with 8-pin PCIe wiring, as those also just need sense pins shorted to ground. But I have never seen this done by anyone, ever in real life. Why? This is speculation, of course, but it's relatively safe to assume that when you're already making a harness of 6 wires, adding two more identical wires is a low-effort undertaking, and coming up with and implementing a shortcut would likely take far more effort than it would be worth. It might even look worse in the end, as 8 matched wires likely look better than 6 matched wires + a couple of bodge wires in the plugs. And, as has been central to my argument all along: these people care mainly about aesthetics.

For these cables, on the other hand, there is no real option for matched cabling (unless, as I tried bringing up as a possibility earlier that you apparently misread, you could find pins for the small connector capable of holding wires of the same gauge as the power wiring - this would let you run 16 equally thick wires for aesthetic purposes, at the cost of unnecessary wiring, exotic and possibly nonexistent pins, and a lot of hassle). You'd then need to find a way to make having four much thinner wires look good alongside 12 thicker ones. How do you do that? Creative routing and combs might help, but that is labor intensive and increases QC costs, which is anathema to these actors. Two-conductor wire of rougly the same thickness as the thicker wire might work, but terminating them in a good-looking way would be a challenge. Or you could stuff two wires into one sleeve - though that would be lumpy and stiff, sub-optimal for aesthetics. The same goes for sticking single wires into oversized sleeves - plus that they would be loose, of course. The third, and suddenly more attractive option, then becomes shorting the wires to ground at the connector. This would be the simplest, cheapest (you're now saving four lengths of wire per cable, rather than the potential two on an 8-pin), and possibly least bodged-together-looking solution of the bunch. That is a problem, when the worst implementation possible has the most advantages.

............................................. I was comparing the pros and cons of going with 16 equally sized pins vs. the solution they actually went for. Was that difficult to understand? Literally the entire basis of my issue here is based on these being different, smaller pins, ffs. I would really suggest you take a step back here, as it seems you're approaching this discussion dead-set on just not understanding what I'm saying, whether consciously or not. I'm sorry if my style of writing isn't sufficiently clear, but literally nothing in the sentence you quoted above is me saying that the pins on the standard presented here are full size.

And I'd agree with that assessment. But I'm also realistic enough to know that people will still do so, and that people fall victim to scammers passing off $5 cables for $100 cables, but with an impressive rebate, etc. That's why standards should prioritize safety to the degree that eliminating unscrupulous shortcut implementation possibilities should be a main goal of the design. That's how you make a safe standard - by not assuming that everyone implementing it is a good actor interested in following it. Instead, what we have here is a standard that, though unintentionally, at least partly incentivizes shortcuts.

boy you assume alot!
here you go, a jumpered 6/8 pin you've never seen:

on a cheap budget abomination of a power supply you assumed i don't know of. it was not uncommon for it to be on a "budget" models sooo the cost is in reality lower, another assumption out the window.

it would be miraculous to be building PCs since college in '94 and not find crap/junk parts. same thing walking into an autoparts or hardware store to find parts for cars, washer/dryers, furnace, hot water tank . . i can buy name brand parts that are on the manufactures' suggested list or go with a no name cheap part and try my luck. stating that there are crappy parts/cables/whatever everywhere (practically since the beginning of time) isn't any excuse as you resentfully put it; it's a statement of fact.

imaginatively find flaws and ranting about the the new cabling on a forum's news section isn't going to change that. going into forum's power supplies sub section and discussing which cables are good and which to avoid (yeah ignoring the bad actors!) is by far much more productive. just like how i found out how bad of a choice i made with the CM PSU - i went to a credible source and found the information to make a better choice.

since you're posting a lot baseless assumptions i'l let you continue making an ass of yourself but i refuse to further make an ass of myself. in the meantime, if you really want to find out and become better informed than i highly suggest you reply to @TheLostSwede 's post to you.

bub bye.

 

[Valantar]

boy you assume alot!
here you go, a jumpered 6/8 pin you've never seen:

on a cheap budget abomination of a power supply you assumed i don't know of. it was not uncommon for it to be on a "budget" models sooo the cost is in reality lower, another assumption out the window.

I have literally never made any statements about your experience with hardware. I asked the rhetorical question (and answer) of "Confidence from knowledge of and experience with what? Seemingly not the cheap custom PSU wiring market, given the approach here" based on the implications of you arguing as if ignoring bad actors is a sensible practice when designing standards. That is not an assumption about your experience, it is my take on how you are arguing here, and the beliefs and ideas embedded in those arguments. I'm sorry if that comes off as an assumption about your experience - it isn't, and was never intended as such.

Also, that is really fascinating - I'm thankfully lucky enough to never have seen a PSU like that.

it would be miraculous to be building PCs since college in '94 and not find crap/junk parts. same thing walking into an autoparts or hardware store to find parts for cars, washer/dryers, furnace, hot water tank . . i can buy name brand parts that are on the manufactures' suggested list or go with a no name cheap part and try my luck. stating that there are crappy parts/cables/whatever everywhere (practically since the beginning of time) isn't any excuse as you resentfully put it; it's a statement of fact.

Again: I've never said anything even remotely amounting to you not having experience with budget/low quality PC parts, I've only commented on the implications of your lines of reasoning and argumentation in this thread.

imaginatively find flaws and ranting about the the new cabling on a forum's news section isn't going to change that. going into forum's power supplies sub section and discussing which cables are good and which to avoid (yeah ignoring the bad actors!) is by far much more productive. just like how i found out how bad of a choice i made with the CM PSU - i went to a credible source and found the information to make a better choice.

Have I ever claimed that me discussing this here is changing anything? How about just ... discussing a potential problem? Is that bad?

But again, you're missing my point: seeking out advice on forums full of well-informed people is not something you can expect even a modest amount of people to do - forum culture is rather and not necessarily the most welcoming, not everyone gets it, and not everyone has the time, energy, and patience to sift through a forum to find answers, or even the base knowledge to understand the answers they get. The implication of this line of reasoning, while obviously never formulated this strongly in your arguments, is that people unable to do this work and gain this knowledge for themselves deserve what's coming to them in terms of poor product quality and safety. (To be clear: you're effectively arguing that people should educate themselves, which begs the question of what happens to those who don't, and on whose shoulders the responsibility for those outcomes fall. Such an individualizing line of reasoning - each is responsible for their own actions, and needs to educate themselves - necessarily leads to those unable to do so also being responsible for any issues.) And IMO, that is exactly the kind of issues that standards are meant to eliminate to the degree possible - they should be aiming for idiot proofing, not "moderately knowledgeable-proofing".

As for the rest of the assumptions I'm apparently making here, I sadly have no idea what you're referring to. I think you might be mistaking my extending of your lines of argument to what I see as their logical conclusions (in order to demonstrate their flaws) as assuming you agree with these conclusions - and to be clear, I don't. People hold opinions that align with things they disagree with all the time - I definitely do as well. But IMO, that's what discussions are good for - highlighting one's own blind spots and the implications of beliefs that one might otherwise not have thought of.

its not only an assumption of my experiences but also inaccurate of what i said. (you're going from point A to point D w/o seeing if B and C are correct.)

I'm well aware of skipping over parts - the entire point was tying together two disparate-seeming parts of your argumentation: claims to experience on the one hand, and a lissez-faire attitude towards safety standards on the other. Those were present, and I put them together to highlight the contrast between the two. Such a move obviously requires speculation as it is literally impossible for me to know what lies between unless you tell me, but itshould be equally obvious that it is a rhetorical device and not actually meant as any attempt at a factually accurate description of your overall attitudes, but simply a comparison of two conflicting parts of those.

not one word about designing practises there or anywhere. i said ignoring bad actors was to not buy their products, had nothing related to design.

But that is precisely my point: by saying that the way to avoid bad actors is to not buy their products, rather than the most obvious solution in this context - to develop standards to make it more difficult to be a bad actor in the first place - you are saying something about design practices. You are expressing an attitude towards individualization of responsibility, as well as attribution of responsibility towards users rather than designers. I disagree with a lot of this, hence my arguments above.

had i said anything, it would that it's been too long since pci-e sig has updated the spec for high powered cards (2008). the limit is 300 watts; pcie, 6 pin and 8 pin. cards with two or more 8 pin power connectors do not comply and cannot use the psi-e sig logo(cert) on the box. this is a very bad optics that products do not need to have standards/certification and needs addressed immediately.

I wouldn't disagree with that, though the important part IMO is that they actually follow the parts of the standard that pertain to safe operations. One could of course argue that a >300W AIC is inherently less safe than a < 300W one, but that's quite a different question, and a rather arbitrary line at that.

discussing a potential problem, no. but discussing FUD, yes. - rhetorically speaking. well and having a discussion about a discussion is . .meh too.

But discussing a discussion becomes necessary once it's clear that the parties involved aren't actually communicating. Which is the case here. As for this being FUD - if you see it as such, you're reading way too much into this. I'm not saying this is a massive problem - I've only been continuing my argumentation in the face of denial of it being a problem at all. I don't foresee this causing issues with serious PSU manufacturers or other large-scale implementations. My point was highly specific and limited in scope. If that's FUD to you, well, that's on you, not me.

well i highly disagree with your assessment, though forums seemly seem toxic these days, its a lot of where you spend your time at. news sub forums are usually the most toxic whereas more specialty - ie power supplies, benchmarking, water cooling, general build help, not as much.

Question: How many people, globally, are even occasionally active on forums pertaining to the things they interact significantly with each day? Are you active on ... dishwasher forums? Public transportation/car forums? Clothing forums? Housing and building maintenance forums?

This is my point: the "people need to educate themselves and take responsibility for their actions" argument is fundamentally utopian and practically impossible. This makes it unreasonable. Nobody on earth has the time and resources to educate themselves meaningfully on everything they interact with, even their main hobbies and practices. This is impossible. People need to be extremely selective about these things. Educating yourself about the things you interact with is of course highly important, but doing so about everything is impossible. This means that in practice, this line of argumentation becomes a cop-out in service of denying responsibility for designers and standards makers. Their job is literally to ensure the safe operation of these things. Not for well-read enthusiasts, but for as many people as possible.

the bottom line is everyone has the duty to understand the potential hazards and safety features before using any device, product, or thing. those that cannot be responsible cause accidents through their own negligence.

Please see above. This is wildly utopian quasi-libertarian ideological nonsense that has zero applicability in real life. This is a standard that is literally impossible for anyone to meet. (Though the wealthy have the advantage of being able to pay others to apply their expertise and education in their stead. Meaning that this line of argumentation is not only utopian, but also inherently classist and discriminatory, as it designates those able to "sensibly" pay others for their expertise as more "responsible", and as such better.)

though i could give a very utopian picture

You already did.

of ethics and moral obligations, (everyone work together!)

... like what a standards body does?

for what would be needed to answer your begging question would be to highly suggest reading about torts

Sorry, I have very little interest in educating myself on the specifics of torts in the US legal system, seeing as it is a deeply, deeply messed up system dominated by frivolous lawsuits and the overall adoption of an absurd litigious attitude, and hardly a standard by which anything ought to be measured.

it is by far not unreasonable ask for anything "moderate." it is unreasonable to ask that standards features

You apparently forgot to finish this sentence, but I get what you're going for. Again, see above: asking that everyone doing something relatively simple (like seeing lots of cool-looking PCs with braided cabling, wanting that for their own, and going searching for cheap cable extensions) educate themselves even "moderately" on something quite technical like PSU wiring safety, to the degree that they are able to identify shady-looking cabling and too-good-to-be-true deals? That is completely unreasonable. What is not unreasonable is to ask a standards body literally comprised on companies consisting of hundreds if not thousands of experts on this field to design a standard that takes as many precautions as possible towards idiot-proofing their standard. And in this case, they instead designed a standard that incentivizes shortcuts.

 

[looniam]

I have literally never made any statements about your experience with hardware. I asked the rhetorical question (and answer) of "Confidence from knowledge of and experience with what? Seemingly not the cheap custom PSU wiring market, given the approach here" based on the implications of you arguing as if ignoring bad actors is a sensible practice when designing standards. That is not an assumption about your experience, it is my take on how you are arguing here, and the beliefs and ideas embedded in those arguments. I'm sorry if that comes off as an assumption about your experience - it isn't, and was never intended as such.

Also, that is really fascinating - I'm thankfully lucky enough to never have seen a PSU like that.


Again: I've never said anything even remotely amounting to you not having experience with budget/low quality PC parts, I've only commented on the implications of your lines of reasoning and argumentation in this thread.

Have I ever claimed that me discussing this here is changing anything? How about just ... discussing a potential problem? Is that bad?

But again, you're missing my point: seeking out advice on forums full of well-informed people is not something you can expect even a modest amount of people to do - forum culture is rather and not necessarily the most welcoming, not everyone gets it, and not everyone has the time, energy, and patience to sift through a forum to find answers, or even the base knowledge to understand the answers they get. The implication of this line of reasoning, while obviously never formulated this strongly in your arguments, is that people unable to do this work and gain this knowledge for themselves deserve what's coming to them in terms of poor product quality and safety. (To be clear: you're effectively arguing that people should educate themselves, which begs the question of what happens to those who don't, and on whose shoulders the responsibility for those outcomes fall. Such an individualizing line of reasoning - each is responsible for their own actions, and needs to educate themselves - necessarily leads to those unable to do so also being responsible for any issues.) And IMO, that is exactly the kind of issues that standards are meant to eliminate to the degree possible - they should be aiming for idiot proofing, not "moderately knowledgeable-proofing".

As for the rest of the assumptions I'm apparently making here, I sadly have no idea what you're referring to. I think you might be mistaking my extending of your lines of argument to what I see as their logical conclusions (in order to demonstrate their flaws) as assuming you agree with these conclusions - and to be clear, I don't. People hold opinions that align with things they disagree with all the time - I definitely do as well. But IMO, that's what discussions are good for - highlighting one's own blind spots and the implications of beliefs that one might otherwise not have thought of.


I'm well aware of skipping over parts - the entire point was tying together two disparate-seeming parts of your argumentation: claims to experience on the one hand, and a lissez-faire attitude towards safety standards on the other. Those were present, and I put them together to highlight the contrast between the two. Such a move obviously requires speculation as it is literally impossible for me to know what lies between unless you tell me, but itshould be equally obvious that it is a rhetorical device and not actually meant as any attempt at a factually accurate description of your overall attitudes, but simply a comparison of two conflicting parts of those.

But that is precisely my point: by saying that the way to avoid bad actors is to not buy their products, rather than the most obvious solution in this context - to develop standards to make it more difficult to be a bad actor in the first place - you are saying something about design practices. You are expressing an attitude towards individualization of responsibility, as well as attribution of responsibility towards users rather than designers. I disagree with a lot of this, hence my arguments above.

I wouldn't disagree with that, though the important part IMO is that they actually follow the parts of the standard that pertain to safe operations. One could of course argue that a >300W AIC is inherently less safe than a < 300W one, but that's quite a different question, and a rather arbitrary line at that.

But discussing a discussion becomes necessary once it's clear that the parties involved aren't actually communicating. Which is the case here. As for this being FUD - if you see it as such, you're reading way too much into this. I'm not saying this is a massive problem - I've only been continuing my argumentation in the face of denial of it being a problem at all. I don't foresee this causing issues with serious PSU manufacturers or other large-scale implementations. My point was highly specific and limited in scope. If that's FUD to you, well, that's on you, not me.

Question: How many people, globally, are even occasionally active on forums pertaining to the things they interact significantly with each day? Are you active on ... dishwasher forums? Public transportation/car forums? Clothing forums? Housing and building maintenance forums?

This is my point: the "people need to educate themselves and take responsibility for their actions" argument is fundamentally utopian and practically impossible. This makes it unreasonable. Nobody on earth has the time and resources to educate themselves meaningfully on everything they interact with, even their main hobbies and practices. This is impossible. People need to be extremely selective about these things. Educating yourself about the things you interact with is of course highly important, but doing so about everything is impossible. This means that in practice, this line of argumentation becomes a cop-out in service of denying responsibility for designers and standards makers. Their job is literally to ensure the safe operation of these things. Not for well-read enthusiasts, but for as many people as possible.

Please see above. This is wildly utopian quasi-libertarian ideological nonsense that has zero applicability in real life. This is a standard that is literally impossible for anyone to meet. (Though the wealthy have the advantage of being able to pay others to apply their expertise and education in their stead. Meaning that this line of argumentation is not only utopian, but also inherently classist and discriminatory, as it designates those able to "sensibly" pay others for their expertise as more "responsible", and as such better.)

You already did.

... like what a standards body does?

Sorry, I have very little interest in educating myself on the specifics of torts in the US legal system, seeing as it is a deeply, deeply messed up system dominated by frivolous lawsuits and the overall adoption of an absurd litigious attitude, and hardly a standard by which anything ought to be measured.

You apparently forgot to finish this sentence, but I get what you're going for. Again, see above: asking that everyone doing something relatively simple (like seeing lots of cool-looking PCs with braided cabling, wanting that for their own, and going searching for cheap cable extensions) educate themselves even "moderately" on something quite technical like PSU wiring safety, to the degree that they are able to identify shady-looking cabling and too-good-to-be-true deals? That is completely unreasonable. What is not unreasonable is to ask a standards body literally comprised on companies consisting of hundreds if not thousands of experts on this field to design a standard that takes as many precautions as possible towards idiot-proofing their standard. And in this case, they instead designed a standard that incentivizes shortcuts.

wow. so i deleted my last reply but you caught it and then found you couldn't post it because i deleted it, so you edited your previous comment and while also changing the context of what i replied to along with editing my deleted reply itself.

you're an F'n piece of work man.

now go bugger off..

 

[Jhart1228]

I don't have a problem with this run away power gobbling - if George Foreman can make my next computer case.

 

[95Viper]

Threads that should/will be closed
Threads of malicious, spiteful or toxic intent against any user, group or company.
This may include a thread going off topic with posts that include the above depending on how the community is responding.