jose2016:

If they make the power supply simpler but the motherboard more complicated we lose the users. I hope that won't happen.

People will upgrade less frequently due to cost. Motherboard prices are already getting ridiculous. This will just drive them even higher.

Making the motherboard be the "power supply" is completely asinine. Anybody actually read when you said it would make setup look cleaner? You would have power cables coming out motherboard to power other parts, like wtf. "Power supply" dies, gonna have to replace motherboard. Also adding this complexity to already a stuffed board and more cost, could also be more fatal to other parts if something goes wrong.

wavetrex:

Pretty much all PSUs today first convert 100-240v AC to 12V DC, then obtain 5V, 3.3V with DC-DC converters inside said PSU. But to be honest this is a good move, Laptops have been powered by 19V DC (most of them) for quite a long time, and they do every other conversion internally ( after the battery regulation ). If these things fit in a slim case of a laptop, it won't be an issue whatsoever to do so on a desktop-sized computer. ... the question is, how long it will take to actually be put in practice. AT to ATX took many years to happen, I'm guessing ATX to 12VO will be even longer. --- p.s. You know what I would find interesting ? To devise a standard for a PSU with a connector in which you can plug a lithium battery pack (also standardized). I'm fine with only high-end PSUs having it... It would make normal desktop computers behave like laptops, just buy a battery pack and attach to it, internal battery backup for power outages, brownouts, spikes etc. No fear of unexpected shutdowns and data corruption... Also, the powered "standby" mode would work nice, keeping data in RAM for Instant-On, even if the computer is unplugged... Yes of course, there are external UPSs, but those are inefficient, heavy, loud, have that giant transformer... as they have to deal with the high AC voltage. This PSU-based battery solution would be slim and elegant, only working at DC levels (12V), and a 13.4V LiPO battery pack can be small and theoretically cheap... as they already are in billions of laptops and other mobile devices. One can dream...

They fit in a laptop because of limited power consumption. You would of course need to scale the size dramatically to service a desktop... thus defeating the purpose and enduing up with yet another voltage regulator.

EspHack:

ive imagined going full DC after eventually getting enough solar panels and batteries, so like 12v wall outlets, and wondered about what to do with my pc and its complicated PSU, now i see thats probably getting addressed later on, and eventually everything will run on usbc, that means almost every device will accept 12v power through it, pretty cool, no more dc>ac>dc

Imagine the size of the conductors you need to run your entire house on 12v DC. There's a very simple reason we use higher voltages.

HeavyHemi:

Imagine the size of the conductors you need to run your entire house on 12v DC. There's a very simple reason we use higher voltages.

This with a caveat, DC can also be higher voltage. A danger with high voltage DC is electrocution makes muscles clamp whereas AC makes you oscillate. AC allows you to release hold of something electrocuting you before becoming crispy. It might not save you but there is a chance.

HeavyHemi:

There's a very simple reason we use higher voltages.

and AC

Vtech:

I have seen this all ready on prebuilt from Fujitsu and Dell, maybe this is not new at all, just not a standard.

That's what i was going to post as well. Or at least, something similar to this. I have had to deal with motherboards that have non-standard power connections and everything such as SSDs get powered by the motherboard. However, i don't really know if the power supplies only had 12v or not.... Such as this motherboard (you can see on the bottom of the motherboard the power connector for SATA HDD/SSD) https://i.ebayimg.com/images/g/05cAAOSwWbVduGdW/s-l640.jpg

Well, there goes the cable management. Now for the price. I wonder how much those PSUs will reduce prices and how much more people will have to pay for MBs. Will it end up being cheaper as sum? Do I need cheaper PSU or MB? Because when I look at intel, I can clearly state that their user base change MBs more often than PSUs while upgrading. So, would I be happy to get $50 cheaper PSU which I replace in 8 years, while I would pay $30 more for each of 3 MBs over same period of time? Not that few $ would phase me. But it really rises question if this was well thought through. And then this will affect all the PC cases manufacturing. Old way of nice cable management which hides cables on route from PSU to final device will be gone. Because then, you look at freaking Pinhead MB. And all those old hiding spaces will become mostly useless waste of space. I wonder what will build of the year look like once everything adapts to new standard. Sadly:

For ATX Form Factor with 12 V connector power supplies.Mechanical dimension of power supplies has not changed from Multi Rail Desktop Power Supplies. Therefore, no need to change the chassis. Following is the current specification

They practically left all PSU sizes same. Compatibility is understandable. But then Big Ass ATX PSU will become another waste of space. No need to change? I am pretty sure that lower PSU complexity will lead to more powerful SFX/TFX/... and that more cases will be built for them. This will change a lot of things over time.

I love the idea of the ATX12VO standard! Why not use a higher voltage though, like 20V or 24V? Would it not enable even less pins and more efficient VRMs? And the option to add a battery connector would be the best thing since canned bread.

Astyanax:

and AC

No, and not AC. Look up Path 65 and educate yourself on transmission of large amounts of current. My point was simply about current. AC or DC is irrelevant. You might want to brush up on your basic Ohms law.

Hopefully they make adapters for current PSUs. Not interested in replacing a good 160 euros PSU because of a 10 cent connector change.

Mufflore:

This with a caveat, DC can also be higher voltage. A danger with high voltage DC is electrocution makes muscles clamp whereas AC makes you oscillate. AC allows you to release hold of something electrocuting you before becoming crispy. It might not save you but there is a chance.

Er no, that is completely false. What kills you is current, not voltage. You can get zopped with 30K volts DC from your spark plug and be fine and get killed by grounding yourself with 120v AC and a fraction of an amp of CURRENT. Current and duration is what kills not voltage. https://www.brighthubengineering.com/power-plants/89792-ac-and-dc-shock-comparison/ Though both AC and DC currents and shock are lethal, more DC current is required to have the same effect as AC current. For example, if you are being electrocuted or shocked 0.5 to 1.5 milliamps of AC 60 Hz current is required and up to 4 mA of DC current is required. For the let-go threshold in AC a current of 3 to 22 mA is required against 15 to 88 of DC current. Facts about Electric Shock It is the magnitude of current and the time duration that produces effect. That means a low value current for a long duration can also be fatal. The safe current/time limit for a victim to survive at 500mA is 0.2 seconds and at 50 mA is 2 seconds.

HeavyHemi:

Er no, that is completely false. What kills you is current, not voltage. You can get zopped with 30K volts DC from your spark plug and be fine and get killed by grounding yourself with 120v AC and a fraction of an amp of CURRENT. Current and duration is what kills not voltage. https://www.brighthubengineering.com/power-plants/89792-ac-and-dc-shock-comparison/ Though both AC and DC currents and shock are lethal, more DC current is required to have the same effect as AC current. For example, if you are being electrocuted or shocked 0.5 to 1.5 milliamps of AC 60 Hz current is required and up to 4 mA of DC current is required. For the let-go threshold in AC a current of 3 to 22 mA is required against 15 to 88 of DC current. Facts about Electric Shock It is the magnitude of current and the time duration that produces effect. That means a low value current for a long duration can also be fatal. The safe current/time limit for a victim to survive at 500mA is 0.2 seconds and at 50 mA is 2 seconds.

Where did I state different? No values were given for use inside a PC, it is a valid point to help determine the maximum voltage that is safe. AC is unlikely to be used inside a PC. When resistance is constant what determines how much current flows is voltage. One doesnt exist without the other. The source has to be capable of enough current flow, but given the context of the conversation, thats a given. If the current flow causes you to grip the voltage source (as is possible with DC) then you will endure current flow for longer than if the source were AC and will not be able to release if the voltage (ie current flow) is high enough. Would you care to point out exactly which parts of what I said are completely false? ps note, I was trying to add to what you said, not oppose it!

HeavyHemi:

Er no, that is completely false. What kills you is current, not voltage. You can get zopped with 30K volts DC from your spark plug and be fine and get killed by grounding yourself with 120v AC and a fraction of an amp of CURRENT. Current and duration is what kills not voltage. https://www.brighthubengineering.com/power-plants/89792-ac-and-dc-shock-comparison/ Though both AC and DC currents and shock are lethal, more DC current is required to have the same effect as AC current. For example, if you are being electrocuted or shocked 0.5 to 1.5 milliamps of AC 60 Hz current is required and up to 4 mA of DC current is required. For the let-go threshold in AC a current of 3 to 22 mA is required against 15 to 88 of DC current. Facts about Electric Shock It is the magnitude of current and the time duration that produces effect. That means a low value current for a long duration can also be fatal. The safe current/time limit for a victim to survive at 500mA is 0.2 seconds and at 50 mA is 2 seconds.

I'll add little something. 12V AC already penetrates your skin enough to allow you to be shocked a bit/or bit more. For DC it is like 20~24V. Bad part is that you may not feel it in contrast to AC. = = = = As for practical side of desire for 20V mentioned in this thread. Change of voltage would require change of all connectors on all devices. As last thing anyone wants is to have 12V plug on their GPU compatible with 20V plug on new generation of cards. Then, all devices would have to switch too. So no upgrades, just getting most of expensive components new. Generally bad idea.

Fox2232:

Well, there goes the cable management.

I could actually imagine this potentially changing the standards for SATA and possibly some other things in a way that cable management would be better. For instance, if all the power is going to the motherboard, why not have the power and data go through the same cable? There'd then be one cable going from the motherboard to a HDD for instance. Probably wishful thinking but honestly this is the only real reason i would say this would be useful.

Aura89:

For instance, if all the power is going to the motherboard, why not have the power and data go through the same cable? There'd then be one cable going from the motherboard to a HDD for instance.

EFI concerns.

Astyanax:

EFI concerns.

Seperation and shielding within a single "cable" is a thing.

Aura89:

Seperation and shielding within a single "cable" is a thing.

True, but it makes cables more expensive. If goal is to make cheaper PSUs, then wasting extra cash on multiple more expensive MBs is reducing given cost benefit of this standard. And if cables have to become more expensive... Other thing is that new cable would require new SATA connector (It is kind of overdue anyway), but that would make all older HDDs unusable on new MBs. Taking all things into consideration, adopting new PSU standard may be beneficial in long run, but will be pretty costly once you do your 1st upgrade. (I am currently happy with 2 reasonably fast M.2 drives and two SSDs:128 + 240GB each dedicated for different OS.) But anyone who is sitting on multiple 1~2TB SATA3 SSDs will not be happy about incompatibility if board does not have enough (or any) legacy SATA3 ports. Times of big technology changes are always costly.

Fox2232:

True, but it makes cables more expensive.

More expensive then separate power and separate data cable? If this ATX12V0 becomes a thing, you'll need a power cable (since it won't come off your power supply) and a data cable. Can't really see that being more expensive then a data + power cable. Only time would tell though. Probably not the best example but there are USB C PD cables that can do up to 100 watts and do data transfer for only $8 a cable. Compare that to a USB cable + Some external power source costs. And no i'm not saying USB is perfect at this, i'm certain there are issues, i'm just saying it's possible without being expensive and i really think overall it'd be less expensive then having to buy a power cable + data cable which is what you'd likely have to do unless PSU's, even non-modular ones, started coming with many cables not attached.

Fox2232:

But anyone who is sitting on multiple 1~2TB SATA3 SSDs will not be happy about incompatibility if board does not have enough (or any) legacy SATA3 ports.

I also don't see where there wouldn't be converting cables. If it was still SATA3 (or backwards compatible) there's no reason there wouldn't be cables that terminate the power and end with a standard (current) data connection or split on the other end to a current standard data connection and power cable. But you're right that it would be a fairly big change, but big changes happen, that's part of what technology is. Personally if this were to become a thing i'd hope that U.2 would become more common, as i understand it anyways, it's already doing what i suggest would happen. U.2, again unless i'm wrong, uses a single cable from the motherboard to the drive, providing PCI-Express lanes similar to M.2, and power, all in one cable. U.2 cables are expensive right now so it may not be the best example again, but that also may simply be due to popularity of the standard and for some reason every one i find has a SATA power cable on it for some reason? Edit: Nevermind, i see why the sata power is there, apparently it needs additional power, which is disappointing....especially since NVME drives get enough power from the M.2 connection. Point being though, since M.2 Drives get power from the motherboard, there should be no major hurdle to make it so it's all part of a U.2-like connection.

12v to the mobo, 12v to the GPU, 12v to a SATA backplane for harddisks, only when necessary. M.2 drives for fast storage, doesn't need power cables. SATA for SSDs is already on it's way into the history books... LED controllers and such can easily be powered by 12v, since they contain PWM DC-DC already inside them... Everything else USB. This is the future of computers for sure.