1080ti having speeds of a 1080.

[CursedGPU10000]

Hi, so I'm wondering whether anyone knows why 1 of my 1080 Ti's is performing at simillar speeds to my 1080.

Current rig:
1 x 1080
2 x 1080ti

All on risers powered by PCI (they're all at Bus: 1) - which I heard barely impacts hashcat speed.
32GB RAM, no CPU bottleneck etc.

Example (running KeePass with -O -w4)

Speed.#1.........:    11836 H/s (393.88ms) @ Accel:64 Loops:512 Thr:1024 Vec:1
Speed.#2.........:    9257 H/s (254.27ms) @ Accel:64 Loops:256 Thr:1024 Vec:1
Speed.#3.........:    8528 H/s (393.15ms) @ Accel:64 Loops:512 Thr:1024 Vec:1

Hardware.Mon.#1..: Temp: 68c Fan: 46% Util:100% Core:1898MHz Mem:5005MHz Bus:1
Hardware.Mon.#2..: Temp: 65c Fan: 37% Util:100% Core:1898MHz Mem:5005MHz Bus:1
Hardware.Mon.#3..: Temp: 65c Fan: 47% Util:100% Core:1898MHz Mem:4513MHz Bus:1


As seen, the GPU falling below standards is: Hardware.Mon.#2
I check in NVIDIA-SMI and it seems to be receiving:

|  2  GeForce GTX 108... WDDM  | 00000000:06:00.0  On |                  N/A |
| 38%  65C    P2  149W / 250W |  2131MiB / 11264MiB |    100%      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+


Whereas the 1080ti which is working great with normal speeds is receiving:
|  1  GeForce GTX 108... WDDM  | 00000000:05:00.0 Off |                  N/A |
| 47%  68C    P2  193W / 250W |  1718MiB / 11264MiB |    100%      Default |
|                              |                      |                  N/A |
+-------------------------------+----------------------+----------------------+


Is the GPU wattage draw causing the low speeds as there's for some reason a 44watt reduction?

It's also worth noting:
Sometimes when trying other hash types, the "GPU falling below standards" initially starts for the first couple minutes performing normally, to near idential speeds as my good 1080 ti. But then there's a low speed drop off.
I've also tried various Hashcat versions and that didn't fix any problems.

And finally to add;
D:\hashcat>hashcat64.exe -m 0 -O -w 4 -b
hashcat (v6.0.0) starting in benchmark mode...

CUDA API (CUDA 11.1)
====================
* Device #1: GeForce GTX 1080 Ti, 9326/11264 MB, 28MCU
* Device #2: GeForce GTX 1080 Ti, 9326/11264 MB, 28MCU
* Device #3: GeForce GTX 1080, 6769/8192 MB, 20MCU

OpenCL API (OpenCL 1.2 CUDA 11.1.114) - Platform #1 [NVIDIA Corporation]
========================================================================
* Device #4: GeForce GTX 1080 Ti, skipped
* Device #5: GeForce GTX 1080 Ti, skipped
* Device #6: GeForce GTX 1080, skipped

Benchmark relevant options:
===========================
* --optimized-kernel-enable
* --workload-profile=4

Hashmode: 0 - MD5

Speed.#1.........: 36112.4 MH/s (51.77ms) @ Accel:64 Loops:1024 Thr:1024 Vec:8
Speed.#2.........: 28015.3 MH/s (66.62ms) @ Accel:64 Loops:1024 Thr:1024 Vec:8
Speed.#3.........: 26671.5 MH/s (50.06ms) @ Accel:64 Loops:1024 Thr:1024 Vec:8
Speed.#*.........: 90799.1 MH/s


 
Any help/advice/tips is greatly appreciated.