Quick News

[Dicehunter]

Another Sapphire 9070XT Nitro 12VHPWR connector melts

 

[Damien c]

I honestly don't understand how this is happening because the AMD cards are not drawing the same power as the nGreedia cards.

Is this a case again of a connector not being fully inserted, bad bend on the cables before the connector etc etc??

 

[Dicehunter]

I honestly don't understand how this is happening because the AMD cards are not drawing the same power as the nGreedia cards.

Is this a case again of a connector not being fully inserted, bad bend on the cables before the connector etc etc??

From what I read there's also no load line balancing on the AMD cards that use this connector so regardless of load it too can melt.

 

[NeverBackDown]

It's always been the connector. The Nvidia issues were that and the fact the power limit of the connector was hardly above full power Nvidia top end cards, just exasperated the problem.

 

[Damien c]

Won't be long before there is a class action in america I suspect, if AMD are cards are now burning up as well.

All because manufacturers decide to cheap out and nGreedia do what nGreedia do best "create something new, force it on all and then let others fix all it's problems".

 

[WYP]

From what I read there's also no load line balancing on the AMD cards that use this connector so regardless of load it too can melt.

This is a part of the problem, and honestly, I do not understand why Sapphire didn't do load balancing and consider the connector as three 8-pins. Split the power evenly over the six 12V pins. 2 pins deliver up to 150W each.

That said, the lack of load balancing is only a problem because the standard has such a low safety tolerance.

It's always been the connector. The Nvidia issues were that and the fact the power limit of the connector was hardly above full power Nvidia top end cards, just exasperated the problem.

Agreed. At a minimum, the standard needs to be updated to some form of load balancing on the powered device (in this case the GPU). It's very simple, consider the connector to be three 200W ports with two 12V pins. Then power the GPU like you would a GPU with three 8-pin power connectors, though with a 200W per connector limit instead of 150W. That is the MINIMUM that needs to be done to make things safer.

Ideally, I would like to see a new standard with a higher safety tolerance that also adopts a form of load balancing on the device side.

Won't be long before there is a class action in america I suspect, if AMD are cards are now burning up as well.

All because manufacturers decide to cheap out and nGreedia do what nGreedia do best "create something new, force it on all and then let others fix all it's problems".

Given the current state of America, I do not see a class action against the country's most valuable company going well. That said, I would love to see some accountability.

Nvidia doesn't allow its high-end GPUs to use alternative power connectors, so Nvidia's partners literally can't use traditional connectors, even if they wanted to.