TREX: Can't find nonce with device [ID=2, GPU #2], Epoch 618 requires 6.33 GB of memory? Faulty epoch on firepow?

[lutsifer]

Hello,
I'm getting constantly this error, ONLY on firepow?!
I am sure it is not related to overclock, intensity or a GPU problem (it happens to all 13 of them), it must be something algo related...
I use another rig with 3080 on same algo - no problem whatsoever..
The problem happens around 20 times per 24 hrs, anywhere between 5 and 90 minutes.
The requested epoch is somewhere in the deep future :D Time travel? :)
Ubuntu 18.04

0231222 04:32:32 firopow epoch: 69, block: 89819, period: 89819, diff: 429.50 M
20231222 04:32:36 [ OK ] 1307/1309 - 116.92 MH/s, 25ms ... GPU #2 | 1.14 G
20231222 04:32:37 [ OK ] 1308/1310 - 116.86 MH/s, 24ms ... GPU #8 | 2.80 G
20231222 04:32:52 TREX: Can't find nonce with device [ID=2, GPU #2], Epoch 618 requires 6.33 GB of memory, but device has 5.94 GB...
20231222 04:32:52 WARN: Miner is going to shutdown...
20231222 04:32:52 Main loop finished. Cleaning up resources...
20231222 04:32:52 ApiServer: stopped listening on 0.0.0.0:2020
20231222 04:32:53 T-Rex finished.
20231222 04:32:54 WARN: WATCHDOG: T-Rex does not exist anymore, restarting...
20231222 04:32:55 T-Rex NVIDIA GPU miner v0.26.8 - [Linux]
20231222 04:32:55 r.104788c2d052
20231222 04:32:55
20231222 04:32:55
20231222 04:32:55 NVIDIA Driver v470.182.03
20231222 04:32:55
20231222 04:32:55 + GPU #0: [00:01.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #1: [00:02.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #2: [00:03.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #3: [00:04.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #4: [00:05.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #5: [00:06.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #6: [00:09.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #7: [00:0a.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #8: [00:0b.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #9: [00:0c.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #10: [00:0d.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #11: [00:0e.0|1c07] P106-100, 6080 MB
20231222 04:32:55 + GPU #12: [00:0f.0|1c07] P106-100, 6080 MB

[lutsifer]

Also this error I am getting only using t-rex (other miners doesn't restart), it will be shame to have to use other miners, somehow I get the best results using your software :(

[lutsifer]

NVIDIA-SMI 470.182.03 Driver Version: 470.182.03 CUDA Version: 11.4