PCI-SIG investigates if PCIe Gen5 “12VHPWR” cable adapters pose overcurrent risk
An article on potential 12VHPWR adapter issues was published by Wccftech. The site managed to obtain an email from PCI-SIG organization to its members, rising a concern over the quality of power cable adapters for 12VHPWR connector.
In the following weeks, we are going to see high-end GPUs with TDPs reaching much higher levels than before. This is also the time when power supply makers introduce brand new ATX3.0 compatible designs that have restrictive power excursion specifications and should already be equipped with modern GPU power connectors (although not necessarily).
The 12VHPWR, PCIe Gen5, 12+4-pin or 16-pin power connector is the newest standard for high-power GPUs such as GeForce RTX 3090 Ti, the first model to support this new connector. NVIDIA has been experimenting with new connectors with the introduction of RTX 30 series and its 12-pin connector, supported by all Founders Edition models.
But even the 12-pin power cable is not sufficient anymore because it was based on dual 8-pin adapter that can only deliver 300W of power by ATX 2.0 specs (plus additional 75W power from the PCIe motherboard slot).
The RTX 40 series will need more power, at least the high-end models will. The RTX 4090 has default TGP of 450W with even higher power for custom models. For such TGP, three 8-pin power cables are needed and even the slightest TGP adjustment means that 3×150W power spec is exceeded.
PCI-SIG has informed its member that the organization is aware of potential thermal variance for 12VHPWR adapters. The organization is currently investigating reports of potential overcurrent risk when using such adapters.
Dear PCI-SIG Member,
Please be advised that PCI-SIG has become aware that some implementations of the 12VHPWR connectors and assemblies have demonstrated thermal variance, which could result in safety issues under certain conditions. Although PCI-SIG specifications provide necessary information for interoperability, they do not attempt to encompass all aspects of proper design, relying on numerous industry best-known methods and standard design practices. As the PCI-SIG workgroups include many knowledgeable experts in the field of connector and system design, they will be looking at the information available about this industry issue and assisting in any resolution to whatever extent is appropriate.
As more details emerge, PCI-SIG may provide further updates. In the meantime, we recommend members work closely with their connector vendors and exercise due diligence in using high-power connections, particularly where safety concerns may exist.
The issue has been explained by Wccftech who set up a test with dual 8-pin and triple 8-pin power adapters used for 600W and 450W load respectively. For the dual 8-pin adapter, the load was uniform, but it exceeded the 150W rating by a magnitude of 2. For the triple-8-pin adapter (the same one that ships with RTX 3090Ti) the power was not proportionally distributed, and it exceeded power draw by 88% for the first 8-pin connector.
12VHPWR Connector To 2 x 8-Pin Adapter In 600W Test Load:
- 1 x 8-Pin Connector = 25.4A or 304.8W (2x Increase Over 150W Rating)
- 1 x 8-Pin Connector = 25.1A or 301.2W (2x Increase Over 150W Rating)
12VHPWR Connector To 3 x 8-Pin Adapter In 450W Test Load:
- 1 x 8-Pin Connector = 25.34A or 282.4W (88% Increase Over 150W Rating)
- 1 x 8-Pin Connector = 7.9A or 94.8W (Within 150W Power Rating)
- 1 x 8-Pin Connector = 6.41 or 76.92W (Within 150W Power Rating)
One should note, however, that ATX 3.0 power supplies are not required to feature Gen5 connector, not even those above 450W. This was confirmed by HardwareBusters. What this means is that 12VHPWR adapters may still be required for ATX 3.0 supplies, should for any reason PSU companies not want to equip their products with the new connector.
While PCI-SIG investigates the thermal variance for existing adapters, users are advised to consider ATX 3.0 power supplies with modern PCIe Gen5 12VHPWR connector natively implemented, as no thermal variance is observed here. The organization is to release a report on the issue soon, possibly before NVIDIA launches its new GPUs. Hopefully, this will also be thoroughly investigated by tech reviewers beforehand.