The problem with the male and female connectors of the so-called 12VHPWR standard has been very public and we have dealt with it carefully as this whole matter progressed. Luckily, and as we saw yesterday, the new standard included within PCIe 5.1 already includes a modernization of it, but how can we differentiate both models? Well, in a simple way, this way you can know if you have one and the other at the same time. compare the 12VHPWR vs 12V-2×6.

A problem, a solution. Although they have really taken their time, and it has cost them a bad reputation and a fairly significant amount of burned graphics, it seems that the PCI-SIG together with Intel and NVIDIA have made a decision in this regard to solve the problems. But there are still doubts in this regard, which we will try to clarify in this article.

12VHPWR vs 12V-2×6, full compatibility

And changing the standard to a completely different one is not necessary as such, although certain improvements and optimizations had to be made to guarantee the connectivity and security of the hardware. Therefore, all the parties involved in the redesign, more than that, It’s really an upgradehave taken into account all the reports studied to create the new standard, which, for unspecified reasons, does not have the same name.

The important thing is that you don’t have to change anything, neither in the GPU nor in the cable, because it is something that will arrive new. That is to say, The new graphics will come with the so-called 12V-2×6 in their female connectorand manufacturers will release power supplies and new cables with this standard.

If you already have a GPU with 12VHPWR and a cable compatible with this standard, you will really only have to change the cable if you want greater security, it’s not mandatory. But here we must clarify several things before continuing so as not to incite error or confusion, which is easy given what we have seen.

Name change: from 12VHPWR to 12V-2×6

As we have already seen, 12V-2×6 is based on 12VHPWR, but what are the differences? Well, there are some, but first, something to understand. Both in its male connector and in its female connector, that is, both on the GPU side (female) and on the power supply cables (male), there will be a name change from 12VHPWR to 12V-2×6 from now on. .

But having said that, on the female connector (GPU) side there will not be any type of physical modification.

That is, it is still a common 12VHPWR in factor and form, but The plastic it is made of receives an update that it shares with the male connector.

Both will have to endure until 105 ºC for 7 days under the most demanding tests. It’s not clear what material is being used now, but it must be of better quality to prevent melting.

Electrical modifications to SENSE pins

Also, now in terms of electrical modification, the power bridges they change their signage between 150W and 300W. Previously, to achieve 150W in SENSE 0 and SENSE 1, they were left open so that the energy could flow. Now, with 12V-2×6 at this power SENSE 0 and SENSE 1 are shorted, which if they go to 300W mode to deliver more power, they change to SENSE 0 to ground and SENSE 1 open:

The 150 watt power level now requires the SENSE0/SENSE1 pins to be shorted on the power supply or cable.

When both SENSE0/SENSE1 are not asserted, as defined by their high impedance Open-Open state, the load cannot consume power. This new SENSE0/SENSE1 combination now defines the new 0 watt state, which was not present in the 12VHPWR connector. These updated SENSE0/SENSE1 combinations ensure that an add-in card can only draw power when the power pins and side band pins are properly mated and secured.

The Open-Open condition indicates that the PCB connector is misplaced or missing. These two conditions are now functionally equivalent. A PCIe CEM 5.1 add-in card will explicitly interpret the Open-Open SENSE0/SENSE1 statistic as the 0 watt setting and will not attempt to draw power from the cable.

It is important that the power supply provides these updated SENSE0/SENSE1 encodings to the cable, to ensure compatibility with PCIe CEM 5.1 compatible add-on cards.

This affects the female and is determined by the graphics card, so it is transparent to the PSU and the user, so it does not imply any cable change if you have a 12VHPWR, since it lacks the 4 Sense pins. Definitely, For it to be effective, a graphics card with 12V-2×6 and a compatible PSU will be needed.

If we have a compatible graphics card (this is expected if you buy an RTX 40 from now on) and your source is 12VHPWR, you don’t have to worry about the source, the graphics card, or the cable, full compatibility, but you will have less energy control lacking, as a rule, the 4 pins mentioned and less security.

Longer pins to improve reliability and secure connection

It was one of the most critical points and has been addressed. Now the pegs are longer and more robust. Specifically, the pins PWR and GND are 0.25 mm longer insidefurther deepto strengthen the connection and also, as we saw yesterday, they are thicker along with their cables. Sense pins reduce their inner GAP by 1.5 mm:

The power pins have been lengthened and the side band pins have been shortened on the PCB header to ensure first/last break mating of the power pins.

This mechanical check is present on the 12V-2×6 PCB header component to ensure that the side band pins engage only after the power pins are sufficiently engaged.

This is done to ensure that the contact is correct and safe. If the 12V-2×6 connector is not plugged in properly it will be easier for the GPU to detect where it is. will shut down quickly to prevent hardware damage thanks to its 4 Sense connectors.

Therefore, it is a step forward in all aspects, because they improve the electrical, torsional and safety aspects, in addition to ensuring the connection against bad contact. There is one detail to take into account that almost no one is seeing.

Intel in the standard ATX 3.1 and PCIe 5.1 indicates that cables cannot be bent shortly after being unplugged. This is explained from the point of view of the thermal cycle of the cable. If we have used the PC for a few minutes, hours or days, the cable will have already reached its operating temperature depending on its AWG.

If we are going to unplug it once the PC has been turned off, it is best not to bend it, because it will still be hot and microcracks may occur inside it. Intel recommends waiting for it to cool, and if possible, not bending it too much at its ends. What is sought is that the current does not flow unevenly between the contacts when the maximum load on the cable is reached or exceeded.

How can you tell the two apart at a glance?

The PCI-SIG has marked this change in a subtle way. Within Intel’s ATX 3.1 and PCIe 5.1 specification we will know if we have 12VHPWR or 12V-2×6 through a detail, since the new connectors, both male and female, are marked as “H++“.

The 12VHPWR connector is marked simply as H+, hence the new name for this new 12V-2×6 standard, which we insist and repeat, is fully compatible, as Intel indicates:

The 12VHPWR connector, introduced in the previous revision 2.0 of this document and in the PCIe CEM 5.0 specification, has been deprecated and replaced by the 12V-2×6 connector, as shown in the PCIe CEM 5.1 specification. While the 12V-2×6 connector specification is mechanically identical to the 12VHPWR connector in most respects, multiple upgrades have been incorporated into the newer 12V-2×6 connector to improve its reliability. It will be marked as H++ to distinguish it from the H+ in force until now.

And so far all the news about the 12V-2×6 connector, which we remember is already reaching the market with the RTX 40 Founders Edition and it is more than likely that assemblers will be introducing it into their models soon.