cancel
Showing results for 
Search instead for 
Did you mean: 

RTX 5090 ASUS ROG Astral - pins exceed 9.2 amps

DGM
Level 7

My RTX 5090 ASUS ROG Astral keeps having issues with the power connector, how can I solve this or get support from ASUS?
"GPU Tweak III reports that one or more pins exceed 9.2 amps or is at 0 amps."

I'm unable to resolve it myself and I've tried all the steps in the FAQ: https://rog.asus.com/support/faq/1055195/
ROG-ASTRAL-RTX5090-O32-GAMING

PSU used: Seasonic Prime PX TX-1600W ATX3.0
Latest NVIDIA drivers.

  • The power connector is plugged all the way in
  • It "clicked"
  • There is no visible gap
  • It is not bend horizontally or vertically 

Utilized the following cables with the same result for each:

  • NVIDIA's 12VHWPR to 4x 8-pin
  • Seasonic 12VHWPR to 12VHWPR
  • Be quiet! 12V-2x6 to 12V-2x6

5090-PowerDetection.jpg
5090.gif
GPUZ.png

594 Views
8 REPLIES 8

DGM
Level 7

Photo's of cables inserted fully without bending.NVIDIA's 12VHWPR to 4x 8-pinNVIDIA's 12VHWPR to 4x 8-pinNVIDIA's 12VHWPR to 4x 8-pinNVIDIA's 12VHWPR to 4x 8-pinBeQuiet-12v-2x6BeQuiet-12v-2x6BeQuiet-12v-2x6BeQuiet-12v-2x6

Serendia
Level 8

Hi there! I suspect a cable failure.

Steps i would do:
- Reseat connectors firmly on GPU and PSU.
- Undervolt GPU via ASUS GPU Tweak III and retest.
- Replace with high-quality, thicker-gauge cable.
- Use a clamp multimeter to check if you have sane values on each wire. (https://www.youtube.com/watch?v=IKLE_mlIKxE)
- Use Thermal Grizzly WireView GPU Power Consumption Meter for precise power monitoring. 
- Try another PSU.

Good luck and be safe.


PARTS USED: GPU used: Zotac RTX 5090 solid OC PSU used: Corsair HX1200i (ATX 3.0 - 2022 revision) Corsair cable: see official site Cablemod cable: 12vhpwr to 4x 8pin sleeved clamp meter used: Kaiweets KC604 DISCLAIMER: each measurment can have a tolerance error of up to 0.1A!

Since the @DGM has already tested several cables, there's no point in testing any other cables.
The @DGM is already using a high-quality power supply and corresponding cables from the power supply manufacturer, as well as the included graphics card power adapter for testing purposes.
There's nothing more to test here.
It's definitely due to the contact between the individual pins on the connector/GPU and the connector/PSU side in the respective sockets.

Unfortunately, this connector is extremely sensitive and has been carefully designed by Nvidia, and this problem keeps occurring.
At first, a cable is correctly seated and doesn't cause any problems, and then the next moment, it happens.
The problem is simply that the power-carrying pins behind the socket are reconnected to form a single cable, thus eliminating balancing, as was previously the case with the 30 series.

It's no use buying multiple expensive cables or power supplies only to end up having these problems again. Thermal Grizzly's WireView isn't helpful here either, as it only displays what the TE already sees through Power Director+ or HwInfo. WireView doesn't actively balance the individual power pins, but only reports/displays them.

But there's supposed to be a solution soon. It's from Aqua Computer. They have or will be bringing an active balancer onto the market that addresses this exact issue and should provide a solution.
The device is called Ampinel.

Intel Core i7 13700K / AiO Fractal Design Lumen S36 v2 RGB / Asus Rog Strix Z790-F Gaming WIFI / Corsair Dominator Platinum DDR5-5600 64GB (4x 16GB) / Asus TUF RTX 5070 Ti OC / 4x Samsung 980 pro 1TB / Seasonic Prime GX 850 W Gold / Fractal Design Meshify 2 Lite RGB Black TG Light Tint / Monitor AOC Q27G2S/EU (WQHD)


@inge70 wrote:

But there's supposed to be a solution soon. It's from Aqua Computer. They have or will be bringing an active balancer onto the market that addresses this exact issue and should provide a solution.
The device is called Ampinel.


Yes, I saw the release announcement, big fan of Aqua Computer products (have a water-cooled system with many of their products) and will certainly pick up the AMPINEL, hopefully the active load balancer will address the issue.

I do strongly believe the ASUS product I bought is defective and unusable in a safe manner with default settings, products or configurations. I would prefer to have a BIOS, Firmware or setting from ASUS that removes the risk of the product burning.

Currently using the system with the ASUS software installed causes it to yell at me to stop using it (GPU Tweak III Power Warnings of AMP's going over the Pins) and there is no non-tinkering way to utilize this product safely.

Silent_Scone
Super Moderator

Hi @DGM 

Can I ask what you're running on the system when 9.2A is exceeded? 

Is the power imbalance similar between the different cables? Or is there a much heavier imbalance when using one or the other between the power pins?

 

9800X3D / 6400 CAS 28 / ROG X870 Crosshair / TUF RTX 4090


@Silent_Scone wrote:

Can I ask what you're running on the system when 9.2A is exceeded? 


Starting up and loading a game of Star Citizen or Jurassic World Evolution 2 on 7680x2160 (Samsung G9 57") is sufficient to immediately trigger the warning and AMP usage. The imbalance seems to be the similair on all cables, spiking to all pins on different intervals. 

The Be quiet! 12V-2x6 to 12V-2x6 was new out of the box and today the Seasonic 12V-2x6 to 12V-2x6 arrived and had the same result instantly.

Utilizing the secondary 12V-2x6 port on the PSU makes no difference either.

 

DGM
Level 7

Support case submitted by email form to ASUS.

Could it potentially be a faulty connector on the GPU side?
Do all RTX 5090 ASUS ROG Astral's have the issue of 12VHWPR / 12V-2x6 pins exceeding 9.2 amps?
Does ASUS confirm it is safe to continue using it in it's stock configuration?

Hi @DGM 
If you've already cross-checked the cables on the BeQuiet, I would initiate an RMA for the GPU. Any issues just give me a shout

9800X3D / 6400 CAS 28 / ROG X870 Crosshair / TUF RTX 4090