D
Nvidia GeForce RTX 4090 Reviews
- Thread starter DavidGraham
- Start date
TopSpoiler
Regular
I want to see someone test the electric current balance of each 12v line of the 12VHPWR in various load scenarios..
DegustatoR
Legend
DavidGraham
Veteran
The problem has been finally identified, the power cables for the 4090 come in two flavors: a 300V (the majority) and 150V (the minority), the 150V are the cause of all of the problems, they are not soldered right and they can't handle sustained high load without damage.
wccftech.com
NVIDIA 16-Pin Adapter Comes In Two Flavors, 300V & 150V
NVIDIA's 16-Pin Adapter cable for GeForce RTX 4090 comes in two flavors, the good 300V and the bad 150V which is prone to melting.
wccftech.com
The whole adapter is such an inelegant solution, I still don't understand why they used the 12-pin connector.
DavidGraham
Veteran
More CPU retesting by TPU, this time they used the 5800X3D, and the gains are even higher than the 12900K!
At 4K, the 4090 is 6.8% faster on average with the 5800X3D (it was 6.5% with the 12900K), with several titles showing 10% or greater performance uplift at 4K, such as Age of Empires 4 (15%), Anno 1800 (31%), Civilization IV (12%), Battlefield V (10%), Borderlands 3 (30%), Death Stranding (17%), Far Cry 5 (30%), Far Cry 6 (8%), Forza Horizon 5 (9%), Halo Infinite (19%), Hitman 3 (19%), Red Dead Redemption 2 (12%), Sniper Ghost Warrior (10%), and Spider-Man Remastered (19%).
At 1440p, the gains are 15% (the 12900K achieved 12%), and 1080p shows crazy 18.5% gains (the 12900K achieved 15%).
RTX 4090 & 53 Games: Ryzen 7 5800X vs Ryzen 7 5800X3D Review
How much performance can you gain with Ryzen 7 5800X3D over the non-3DV Cache Ryzen 7 5800X when using the mighty GeForce RTX 4090? Our review has the answer, we test 53 games at three resolutions.
Beware, the original TPU review (with the 5800X) tested only 25 games, these remade CPU reviews (12900K/5800X3D) test with 53 games. So results are not directly comparable, however, 17 of the original 25 games showed statistically significant increases with the 5800X3D, with an average of ~10% uplift at 4K, and about 9.5% uplift with the 12900K.
Last edited:
Surely the 150V or 300V label doesn't determine how the cable is soldered? Both can be soldered properly or improperly, no?
Seems a bit premature to announce this is the issue.
I’d guess it’s two different suppliers and the supplier using the 150V rated wire just has poorer quality control. They probably have some slipping through that have too little solder or cold joints.
TopSpoiler
Regular
300V cable, Pin #4 melted:
DegustatoR
Legend
One would think that if the problem were in cables the cables would be melting, not the pins in 12VHPWR plug.
D
Deleted member 2197
Guest
No FE models reported to have the problem yet? Currently only some AIBs?
Yah, the cable should have a temperature rating for the insulation that will also affect the safe current carrying capacity.
That's probably true. Usually if you push a cable too hard you'll see the insulation around the entire cable will start to get soft and eventually melt or burn off. It actually seems like the connector itself may not be mating perfectly. Any time you have an electrical connection that's loose or intermittent, you'll see that connection point heat up and burn out. I've seen the same thing with much bigger and much scarier connections. One of my coworkers forgot to crimp a connection inside a dc power system at work. He basically put a lug on a wire, didn't crimp it, and then put heat shrink over top so you'd never know. Then he connected it to a 100A breaker lol. Eventually some equipment powered off and when I want to investigate it had burned up completely black. It was probably arcing constantly.
If I had to guess, the solder might be breaking as the cable flexes. Could be touching so there's a connection, but it starts to heat up until it eventually burns out. Something like that.
If the solder would break, the heat would be spreading more to the other pins or just desolder the wire. In the adapter all pins are shorted in between with a relative thick metal.
My guess, because in the adapter all pins are shorted, there is no load balancing happening at the pins anymore, the pin with the least resistance (contact plus pcb) will deliver the most current and heat up till it burns.
Considering there is only a 10% safety margin per pin in the new connector it could be overloaded quite fast.
Buildzoid seems pretty sure it isn't the cable, as the wrong end is melting if it were.The problem has been finally identified, the power cables for the 4090 come in two flavors: a 300V (the majority) and 150V (the minority), the 150V are the cause of all of the problems, they are not soldered right and they can't handle sustained high load without damage.
NVIDIA 16-Pin Adapter Comes In Two Flavors, 300V & 150V
NVIDIA's 16-Pin Adapter cable for GeForce RTX 4090 comes in two flavors, the good 300V and the bad 150V which is prone to melting.
Weren't the first reported burns FE models?
D
Deleted member 2197
Guest
AFAIK only AIB models (Asus, Zotac, Galaxy and some others) have reported the issue. Someone on Reddit keeping tally noted no FE GPU's have reported the cable issue yet.
Microbenchmarks of RTX 4090 at Chips and Cheese: Microbenchmarking Nvidia’s RTX 4090. The tests suggest Nvidia's most significant changes with the Ada architecture can be found in the memory infrastructure to enable Ada to scale up beyond Ampere notwithstanding the lack of appreciable increases in VRAM bandwidth.
On another note, I picked up my 4090 FE today. It's built like a brick (in a good way) and I had to remove the side panel from my Lian Li O-11 Dynamic to accommodate the Nvidia adapter. I'm eagerly awaiting CableMod's 180 Degree adapter. I haven't had much time to play with the card other than to do a quick run of Cyberpunk at Max (Psycho) settings. It's very quiet though I did notice a bit of coil whine when the GPU clocks over 2700mhz or so. At a cursory glance, changing the power limit from 133% to 100% to 80% didn't affect performance very much at all (but did lower power by 100w from min to max). I'm currently running a 1440p 165hz monitor.... could be I'm hitting a CPU bottleneck on my 5900x and/or relatively slow system memory (DDR4 3200 16-18-18-36).
On another note, I picked up my 4090 FE today. It's built like a brick (in a good way) and I had to remove the side panel from my Lian Li O-11 Dynamic to accommodate the Nvidia adapter. I'm eagerly awaiting CableMod's 180 Degree adapter. I haven't had much time to play with the card other than to do a quick run of Cyberpunk at Max (Psycho) settings. It's very quiet though I did notice a bit of coil whine when the GPU clocks over 2700mhz or so. At a cursory glance, changing the power limit from 133% to 100% to 80% didn't affect performance very much at all (but did lower power by 100w from min to max). I'm currently running a 1440p 165hz monitor.... could be I'm hitting a CPU bottleneck on my 5900x and/or relatively slow system memory (DDR4 3200 16-18-18-36).
Henry swagger
Newcomer
Microbenchmarking Nvidia’s RTX 4090
Nvidia’s RTX 4090 features Nvidia’s newest architecture, named Ada Lovelace after a pioneer in early computing. Compared to their previous architecture, Ampere, Ada Lovelace enjoys a pr…
chipsandcheese.com
