Thursday, January 23rd 2020

Pure 12V PSU Standard, Named ATX12VO, Debuts Later This Year

Back at CES, at the FSP booth, we spied an inconspicuous-looking PSU with a curious 10-pin connector in place of the 24-pin ATX. The FSP500-30AKB turned out to be the first public exhibit of a the pure 12-Volt PC power supply standard being pushed by Intel, which is called "ATX12VO," which abbreviates Advanced Technology eXtended 12-Volt Only. According to the specification, the PSU only puts out +12 V and 12 Vsb voltage domains, and does away with the 5 V, 5 Vsb, and 3.3 V domains. This greatly simplifies the design of PSUs, as PCs of today don't use too many power-hungry 5 V or 3.3 V devices (such as half-a-dozen mechanical hard drives). The PC will still need 5 V for interfaces such as USB, but VRM on the motherboard will be responsible for DC-to-DC switching of 12 V to those lower-voltage domains. It's also likely that the motherboard will now put out a handful SATA power connectors.

Intel could debut ATX12VO within 2020 via its next-generation desktop platform, which features a 10-pin connector instead of 24-pin. It remains to be seen if the company could help the transition from current PSUs to the new standard by having its motherboard partners include a 24-pin to 10-pin adapter of some sort. In addition to the 10-pin connector, ATX12VO PSUs will put out two other purely-12 V connector types: 8-pin/4+4 pin EPS and 6+2 pin PCIe power. The EPS connector powers the CPU VRM, while the PCIe connector powers add-on cards, such as graphics cards. 4-pin Molex connectors could also be put out, but those will only feature 12 V pins (the 5 V pins will be absent).
Source: CustomPC
Add your own comment

102 Comments on Pure 12V PSU Standard, Named ATX12VO, Debuts Later This Year

#26
TheLostSwede
Also, it might be worth keeping in mind that we're most likely looking at a transition period of at least 5-6 years for this to happen. Not sure how many of you are old enough to remember the AT standard. It took quite some time for everything to transition from that to ATX, it didn't happen over night, neither will this, so take a chill pill.

Chrispy_
Not to mention that portable mechanical drives powered solely by USB cables have outsold internal SATA-cable drives something like 10:1 for the last few years. USB drives are the overwhelming majority of mechanical drives bought by consumers and internal SATA cable drives are rapidly becoming a niche product. You can already buy dirt-cheap USB to SATA power cables on eBay, Amazon, AliExpress etc.
If nothing else, this shows how easy it is to make good 12V to 5V power converters, as the drive enclosures tend to come with 12V only power adapters. This means the other conversion is done by the small PCB that also contains the USB to SATA chip.
Posted on Reply
#27
londiste
If this takes off, pretty sure a lot of both motherboards and PSUs will support both simultaneously with simple adapters of one sort or another, at least during the transition period.
Posted on Reply
#28
TheLostSwede
ppn
Why the 10 pin, just make it 4 pin for the power and standby voltage, all other 8 pin and 6+2.
Because you don't want to put too high Amperage through a single pin/wire. You can only put so much current through a certain thickness of materials, once you go beyond that, bad things are going to start happening. It's also getting to a point where it's dangerous if you touch it by accident and you could get electrocuted, even from a DC power source.
Splitting things up also allows higher current to reach the device, while keeping things safe per wire, although there's a trade-off here as well to make sure it's as efficient as possible. Then there's things like impedance that could reduce cause some other funny things to happen, such as Voltage reduction, which you don't want either.
Posted on Reply
#29
Valantar
londiste
Pretty sure that in case motherboard manufacturers fail to do power conversion job well there will be small power conversion boxes developed to take care of that problem :p
Silverstone would make them. Silverstone makes everything.
Posted on Reply
#30
iO
Should have happened long time ago.

But 4.5 Watts per USB3 port and 8 Watts per M.2 slot add up pretty fast and I bet the MB manufactures hate the idea of putting a ton of additional VRMs on the boards.
Posted on Reply
#31
TheinsanegamerN
Chrispy_
/Facepalm.

Read the article. This isn't abandoning support for mechanical drives. It's simply changing the location of the 12V-to-5V conversion circuitry. If you buy a new 10-pin PSU and board with SATA ports, it will still have a way to power your SATA drives.

The new PSU standard provides the option of dropping 5V and 3.3V support to board manufacturers, whilst everyone can benefit from smaller, simpler, cheaper, better power supplies. It is 100% advantageous to everyone and has no downsides.



Not to mention that portable mechanical drives powered solely by USB cables have outsold internal SATA-cable drives something like 10:1 for the last few years. USB drives are the overwhelming majority of mechanical drives bought by consumers and internal SATA cable drives are rapidly becoming a niche product. You can already buy dirt-cheap USB to SATA power cables on eBay, Amazon, AliExpress etc.
You know, because motherboard manufacturers always supply enough ports and connectors, they NEVER cheap out on connectors leading end users to needing expansion cards, no siree. I like the idea of a 10 pin connector, but lets not pretend that motherboard makers are going to supply enough connectors for all users without some enforced standard.

iO
Should have happened long time ago.

But 4.5 Watts per USB3 port and 8 Watts per M.2 slot add up pretty fast and I bet the MB manufactures hate the idea of putting a ton of additional VRMs on the boards.
Yeah, where are mobo makers going to put all this stuff? The 5V and 3.3V boards in power supplies are reasonably large, and most full ATX boards dont have huge patches of unused space, let alone micro atx or mini itx boards. Where on earth are they going to put this stuff? Look how big the VRM for the CPU is, even at half that size, thats still a respectable amount of additional circuitry.
Posted on Reply
#32
Ferrum Master
TheinsanegamerN
Look how big the VRM for the CPU is, even at half that size, thats still a respectable amount of additional circuitry.
Why do we have already ITX 12V only offerings... how on earth laptop boards even work having +18V supply... gosh... is it a miracle? :shadedshu:

There are absolutely zero electrical problems. It all just about the habbit.
Posted on Reply
#33
iO
TheinsanegamerN
Yeah, where are mobo makers going to put all this stuff? The 5V and 3.3V boards in power supplies are reasonably large, and most full ATX boards dont have huge patches of unused space, let alone micro atx or mini itx boards. Where on earth are they going to put this stuff? Look how big the VRM for the CPU is, even at half that size, thats still a respectable amount of additional circuitry.
It's certainly doable but it won't be easy and therefore not cheap.
But I guess this has the same fate as the BTX standard: won't gain much traction outside the OEM space and will die eventually...
Posted on Reply
#34
TheLostSwede
iO
Should have happened long time ago.

But 4.5 Watts per USB3 port and 8 Watts per M.2 slot add up pretty fast and I bet the MB manufactures hate the idea of putting a ton of additional VRMs on the boards.
Just to show you what a modern 12V to 5V conversion circuit looks like. What you can't see is a pair of smoothing capacitors on the other side of the PCB.
This is something I've done for a client, very simple, very effective and it can do at least 5V/2A. That PCB is 60x40mm and it's just so it can accommodate enough connectors on the other side. Admittedly this would not be how a motherboard would be designed, as they'd use a more powerful setup that could power all the 5V outputs, but it's hardly as if this is going to be a major issue for the board makers, compared to doing the CPU VRM.



TheinsanegamerN
Yeah, where are mobo makers going to put all this stuff? The 5V and 3.3V boards in power supplies are reasonably large, and most full ATX boards dont have huge patches of unused space, let alone micro atx or mini itx boards. Where on earth are they going to put this stuff? Look how big the VRM for the CPU is, even at half that size, thats still a respectable amount of additional circuitry.
See above, it's a non issue. Also, plenty of mini-ITX boards already in the market with 12V only, either via a barrel plug or a 2/4-pin connector.
As an example, see link below.
https://www.gigabyte.com/Embedded-Computing/GA-IMB310TN-rev-10

Ferrum Master
Why do we have already ITX 12V only offerings... how on earth laptop boards even work having +18V supply... gosh... is it a miracle? :shadedshu:

There are absolutely zero electrical problems. It all just about the habbit.
Actually, most modern notebooks are 19V ;)

iO
It's certainly doable but it won't be easy and therefore not cheap.
But I guess this has the same fate as the BTX standard: won't gain much traction outside the OEM space and will die eventually...
But it is cheap. The parts needed are few cents. I don't get it, why are so many people in this thread making this out to be something really complex, when it's not?
Another project, this is a pico-ITX board at 100x72mm with 12V input that can deliver 5V and 3.3V to the pin headers. How can such a tiny board do it when some people here claim there isn't space on a mini-ITX board?
Posted on Reply
#35
londiste
TheinsanegamerN
Yeah, where are mobo makers going to put all this stuff? The 5V and 3.3V boards in power supplies are reasonably large, and most full ATX boards dont have huge patches of unused space, let alone micro atx or mini itx boards. Where on earth are they going to put this stuff? Look how big the VRM for the CPU is, even at half that size, thats still a respectable amount of additional circuitry.
In the space that is left over from 10-pin connector being smaller than 24-pin one?
I am joking, but it isn't too far from the truth :D
Posted on Reply
#36
windwhirl
TheLostSwede
However, it might lead to us seeing proper backplanes in more cases, which is how SATA should've been designed to work from the beginning anyhow. The wired SATA power cable is simply horrible imho.
You know what? Forget what I said, I'm all for it if it somehow solves my rat nest of SATA cables.
Posted on Reply
#37
Valantar
TheinsanegamerN
Look how big the VRM for the CPU is, even at half that size, thats still a respectable amount of additional circuitry.
...CPU VRMs deliver hundreds of amps. CPUs run at anywhere from 1V to 1.4V (-ish) while consuming ~50-300W. Motherboard VRMs are built so that they can at the very least support the highest end CPU for that platform at stock, meaning that for example any H310 or B350 motherboard VRM is built to deliver power to a 9900K. That means delivering its ~160W PL2 for short term power and ~95W PL1 for long term power without issue - and that applies to even the most lowly board, ignoring the massively overbuilt VRMs for high end overclocking. I don't remember what voltage a 9900K runs at under all-core load, but if we say 1.3V and 160W power consumption, that's 123 Amps. Now, you won't get a 123A 5V VRM in the same space as a 123A 1.3V VRM, but not far from it. And 3A per USB port is generous, meaning that a "half sized" VRM like what you describe could power ... somewhere above 30 USB ports. You can get away with far smaller VRMs than even cheapo CPU VRMs if all you're doing is powering USB ports, in other words.

As for voltage converters, we shouldn't forget the existence of monolithic DC-DC converter blocks like this one from Murata. 5V4A output from 12V in a 9x10.5x5.5mm package. That's essentially the size of a CPU VRM inductor alone. Add it to the board and you can power at least a few USB ports. Sure, this one is expensive and relatively low power, but there are cheaper ones, and they definitely get cheaper if you're buying ten thousand a pop.

Nonetheless, people need to stop thinking this will make motherboards noticeably more expensive. It won't. The components needed are well known to manufacturers - they can even likely reuse CPU VRM components if they want to - compact and readily available. Also, comparing on-board component sizes to PSU component sizes is an invalid comparison, as PSUs are always overbuilt; they are built to cover a wide variety of hardware configurations, while an on-board supply would be tailored to the components on that board. Even looking at a lowly 500W unit like the most recent PSU review here on TPU, it supports 17A on its 5V rail and 18A on its 3.3V rail. Both of those are massively overkill for most modern usage scenarios, and could thus be downsized significantly if moved to the motherboard. A 500W PSU needs to be built to handle both a gaming setup with a fast CPU and GPU and a single SSD and a file server with a low-power CPU, no GPU and a heap of mechanical HDDs and SSDs, meaning they are built for a wide variety of different loads. With this standard we can reduce the overhead here and move towards more tailor-made solutions better suited to their intended use.

Gungar
Higher efficiency? xD you know that you just moved the problem on the motherboard right? you gain efficiency on the psu and you lose on the motherboard there is 0 changes to efficiency xD
Not true. There are several ways in which this improves efficiency. None are dramatic improvements, but they all matter:
  • No cable losses transferring 3.3V and 5V from the PSU to the motherboard (and likely shorter cables for any devices needing power from the motherboard, like HDDs)
  • Simpler, cleaner PSU designs due to only producing a single output voltage (this is the biggest one by far)
  • Universal adoption of DC-DC conversion for lower voltages (doesn't apply to modern high-end PSUs that already do this, but many still don't)
  • Better sizing of DC-DC converters for the load in question, hitting efficiency sweet spots more easily
  • Proliferation of bespoke converters for power-hungry components (
londiste
In the space that is left over from 10-pin connector being smaller than 24-pin one?
I am joking, but it isn't too far from the truth :D
Definitely not far from the truth. Delivering a few amps for USB and SATA really doesn't require a lot of space.

Another pro for this, btw: moving to motherboards needing voltage converters for USB could easily lead to a proliferation of USB-PD support on the same boards - it's not that much more difficult building a 12V-to-5/9/12/15/20V system than a 12V-to-5V one.
Posted on Reply
#38
Ferrum Master
Also it will leave more space in the PSU itself, better ventilation for smaller form factors SFX, where it matters.

Gosh... space? The whole backside of any motherboard is mostly deserted, with some caps under South bridge and CPU.
Posted on Reply
#39
R-T-B
TheLostSwede
I'm surprised this hasn't happened sooner tbh. With storage moving towards M.2, even things like SATA and Molex connectors aren't going to be widely used, apart from when people are using their PC's and keep done older parts.
It's also quite efficient to convert 12V DC to lower DC Voltages, especially compared to going from AC to DC. I've done some power conversion boards for a client and they were quite cheap to make and used very few components, yet can support quite high loads without producing much heat.
I agree it is overdue, but as most m.2 boards are running at 3.3v isn't that kind of a poor example?
Posted on Reply
#40
iO
TheLostSwede
But it is cheap. The parts needed are few cents. I don't get it, why are so many people in this thread making this out to be something really complex, when it's not?
Another project, this is a pico-ITX board at 100x72mm with 12V input that can deliver 5V and 3.3V to the pin headers. How can such a tiny board do it when some people here claim there isn't space on a mini-ITX board?
Fair enough for a SBC but what about an ATX sized board?
Let's say 8 USB ports on the IO and 4 internal each 900mA. Plus 7 PCIe slots with each 3A and two M.2s with 2.5A each. That's over 50 Watts for the 5V and 80 Watts for 3.3V supply.
Putting additional VRMs of that dimensions on an already feature packed PCB is tricky and costly.
Posted on Reply
#41
TheLostSwede
The problem with USB these days is that with USB-C, at least some ports can (should?) be as much as 100W, but that's at 20V/5A. However, it's unlikely we'll see this on any desktop PC, but maybe these new power supplies will make 12V/5A more likely to appear than today.
However, long gone are the days of 500mA or 900mA USB connectors, as most boards deliver 1-2.4A per USB ports these days, simply because people expect to be able to charge their mobile devices from the ports on their PC.

@Valantar That Murata part is insanely expensive though, so that's not likely to find its way into regular PCs as yet. Also, high Amps on the minor rails is a way that the PSU makers have managed to "uprate" their PSUs for years, since it's the combine Wattage that's put on the box at the end of the day...
Posted on Reply
#42
Ferrum Master
iO
Putting additional VRMs of that dimensions on an already feature packed PCB is tricky and costly.
Quite joking Dude :D :D We are actually fed the with the lowest crap grade boards nowadays... tomato boards mostly... and paying great amount of money for them. It adds up NOTHING in design process. It is not a RF part where you design high frequency transmitter, this basically basic grade newbie task to add... copy paste in the PCB design program.

There are PMU's capable of delivering many amperes of current... multi rail, multiple regulation and safety, low noise.

Haven't you thought that we have now wall chargers with GaN? Tiny things with a primary for mains voltage pumping out 60W?? HELLOOO people? What is this?

It amuses me really...

I have a new argument.

How about putting less RGB PWM drivers on the board, how we got place for them then? :laugh:
Posted on Reply
#43
TheLostSwede
R-T-B
I agree it is overdue, but as most m.2 boards are running at 3.3v isn't that kind of a poor example?
It wasn't about M.2 as such, but the fact that new systems are likely to use fewer drives that require power cables from the PSU.

iO
Fair enough for a SBC but what about an ATX sized board?
Let's say 8 USB ports on the IO and 4 internal each 900mA. Plus 7 PCIe slots with each 3A and two M.2s with 2.5A each. That's over 50 Watts for the 5V and 80 Watts for 3.3V supply.
Putting additional VRMs of that dimensions on an already feature packed PCB is tricky and costly.
Why is there a difference? Again, see above, USB ports on most motherboard are no longer 900mA, as it's not enough for charging phones etc. with.
Still, you're going from 12V, so it's not expensive or tricky.
This is a new part from TI that can do 40A and it's a 5x7mm chip. That means it can be placed pretty much anywhere on the motherboard. Yes, this part is a bit on the pricey side, but it was the first thing I found doing a quick search and I'm sure it's far from the only part like this.
http://www.ti.com/product/TPS543C20A
Posted on Reply
#44
Ferrum Master
TheLostSwede
Yes, this part is a bit on the pricey side
Reel costs less than 5€ per unit... a 40A part... that is not pricey...



Dev board size

Posted on Reply
#45
sutyi
hojnikb
Reducing footprint and complexity of a PSU? It's actually great and a long overdue option for PCs.
If you get a decent PSU you'll probably use it 5+ years. Probably swapping motherboards every 2-3 years.

You are suffling the same components from the PSU to the MB and ending up swapping them twice as fast.

Current decent PSU designs are already DC-2-DC. I would rather have all the filtering and protection in the PSU housing, instead of using my brand spanking new motherboard effectively as power distribution panel. You would also need a lot of digital short circuit protection and filtering added to the VRM of the motherboard, while also adding a lot of complexity to the PCB it self, so memory and PCIe traces have proper shielding.

Not sure if I'm ready to buy 200USD entry level motherboards cause they need to be manufactured on 8-10 layer PCBs...
Posted on Reply
#46
TheLostSwede
Ferrum Master
Haven't you thought that we have now wall chargers with GaN? Tiny things with a primary for mains voltage pumping out 60W?? HELLOOO people? What is this?
You just reminded me, I have a friend that works here.
http://www.exagan.com/en/
Posted on Reply
#47
Ferrum Master
sutyi
Probably swapping motherboards every 2-3 years.
Why it you bother about protection etc... actually the solutions FURTHER away from mains give more safety, they are more safe and tailored to their task, as you can cherry pick them for the device it sits after? USB short cut protection is already there but actually driven by a dedicated power IC it will work faster and better giving more safety options.

Also you are in EU? 2 years of warranty... who cares if you change them anyways?

TheLostSwede
You just reminded me, I have a friend that works here.
http://www.exagan.com/en/
It's kinda a hot topic now... I have to take a shovel and take all my chargers to the bin because of that as they all had gone EOL.
Posted on Reply
#48
sutyi
Ferrum Master
Why it you bother about protection etc... actually the solutions FURTHER away from mains give more safety, they are more safe and tailored to their task, as you can cherry pick them for the device it sits after? USB short cut protection is already there but actually driven by a dedicated power IC it will work faster and better giving more safety options.

Also you are in EU? 2 years of warranty... who cares if you change them anyways?



It's kinda a hot topic now... I have to take a shovel and take all my chargers to the bin because of that as they all had gone EOL.
We were talking footprint. Not warranty. If I buy the same components at twice the rate, I have twice the ecological footprint... makes sense? Or is my train of thought wrong over here?
Posted on Reply
#49
Ferrum Master
sutyi
We were talking footprint. Not warranty. If I buy the same components at twice the rate, I have twice the ecological footprint... makes sense? Or is my train of thought wrong over here?
They would be used anyways? Either in the PSU or on the board, with less wires, less plastic, same evil, isn't it? It is up to the manufacturer how fragmented it will be. One is sure... due to USB-PD it will be designed and receive independent power IC anyways... old ATX or the new one. The small rails? Some step down solution IC's are made in such amounts already, they are just idling there... there will be no change regards ecological footprint. Power consumption actually with proper gating, idle features should be lower.
Posted on Reply
#50
TheLostSwede
sutyi
We were talking footprint. Not warranty. If I buy the same components at twice the rate, I have twice the ecological footprint... makes sense? Or is my train of thought wrong over here?
Again, this is NOT going to happen over night. So in your case, you can most likely match the upgrade pattern you use to fit in a new PSU when it suits you.
Posted on Reply
Add your own comment