News Nvidia confident that RTX 50 series power connectors unlikely to melt despite higher TDP

So are they using the obsolete 12VHP or the new standard 12V-2x6, or is it a vendor's choice situation where some use one and others use the other?

Maybe talk to Nvidia and get an official statement?
All 50-series with a 16-pin connector, and most 40-series, will use 12V-2x6. I'm not sure what 40-series GPUs didn't update; I think mostly it was unsold cards from the earlier production runs that still have the 12VHWPR.
 
All 50-series with a 16-pin connector, and most 40-series, will use 12V-2x6. I'm not sure what 40-series GPUs didn't update; I think mostly it was unsold cards from the earlier production runs that still have the 12VHWPR.

The RTX 50-series GPUs, part of Nvidia's next-generation line-up, will reportedly continue to use the PCIe Gen 5 16-pin connector, also known as the 12VHPWR connector.

https://www.tomshardware.com/pc-com...o-feature-yellow-tipped-16-pin-power-adapters

So now "12VHPWR" and "12V-2x6" are going to be used interchangeably? I can't see any way that could possibly get confusing...
 
https://www.tomshardware.com/pc-com...o-feature-yellow-tipped-16-pin-power-adapters

So now "12VHPWR" and "12V-2x6" are going to be used interchangeably? I can't see any way that could possibly get confusing...
There is no visible differences between "12VHPWR" and "12V-2x6", you can use them interchangeably. Manufacturers have just rebranded the products, several without changes. "12VHPWR" is defined by specifications CEM 5.0 and "12V-2x6" by CEM 5.1. The commercial rebranding was just to reassure users.
CEM 5.1 address two causes or errors from the twelve causes of melting analyzed on CEM 5.0, it was the two main causes, it's means that ten causes of melting remains on CEM 5.1.
CEM 5.1 complete the specifications by example in adding on tolerances and alloys, trying to set a minimum in the product quality. It's remains up to manufacturers to ensure the products are up to specifications and safe (they are holes in the specifications remaining to ensure a minimum of surface of contact).
Every month they are several cards or connectors stampeded 12VHPWR melting because of defective connectors that should have not been passed by control of quality. The main culprit for those cases are the square pins which are difficults to produce with accuracy.
 
  • Like
Reactions: Peksha
Still seems like a bad plug to me. Not enough metal per port or something, just freeky.

That said, has there been any word on what initial stock is going to look like? Hoping it will be huge so we actually have a small chance in hell of getting a 5080 or 5090 out the gate.
 
There is no visible differences between "12VHPWR" and "12V-2x6", you can use them interchangeably. Manufacturers have just rebranded the products, several without changes. "12VHPWR" is defined by specifications CEM 5.0 and "12V-2x6" by CEM 5.1. The commercial rebranding was just to reassure users.
CEM 5.1 address two causes or errors from the twelve causes of melting analyzed on CEM 5.0, it was the two main causes, it's means that ten causes of melting remains on CEM 5.1.
CEM 5.1 complete the specifications by example in adding on tolerances and alloys, trying to set a minimum in the product quality. It's remains up to manufacturers to ensure the products are up to specifications and safe (they are holes in the specifications remaining to ensure a minimum of surface of contact).
Every month they are several cards or connectors stampeded 12VHPWR melting because of defective connectors that should have not been passed by control of quality. The main culprit for those cases
are the square pins which are difficults to produce with accuracy.

Corsair made a lovely image which visualizes the difference: 12V-2x6 uses shorter sense pins and longer power pins than 12VHPWR. You can use either cable interchangeably since a complete and firm seating satisfies both, but I would in no circumstance trust any new GPU that still uses the old 12VHPWR connector instead of the 12V-2x6.

12vhpwr_vs_12v2x6_xHQcgFn.width-1000.format-webp.webp
 
  • Like
Reactions: TJ Hooker
Yea, with that still not covered by warranty it gives "full confidence" for a $2k without sclaping price...

At one point I have a feel maybe it's better if they just built in the first 10cm of flexable cable dangling out the card like a tanticle will be a better idea... much easer to make sure that dangling connector is fully seated than with the embedded socket with all sorts of orientations..
 
Nvidia was confident last time as well. And then tried to double down and say it was all only user error. These days, I believe it when I see it, which is why I almost never am an early adopter anymore.
same boat here, especially when the power consumption goes up and the price just sky rocket... it's not like it's something just breaks and won't hurt too much