2023.1.1 (Windows): Out of memory message

[UselessGuru]

09:46:18 Start API Monitor Listener 09:46:18 Start mining 09:46:18 Mining fee for 'ETHASH': 1.00% 09:46:18 Start worker loop... 09:46:18 Pool[DEF-0]: Connecting to: stratum+tcp://daggerhashimoto.auto.nicehash.com:9200 09:46:18 Pool[DEF-0]: Connected to: daggerhashimoto.auto.nicehash.com:9200 - IP4/IP6: 34.149.10.66 09:46:18 Pool[DEF-0]: Set pool extra nonce to: 0x672C82 09:46:18 Pool[DEF-0]: Difficulty changed from 0 H to 1.708 GH 09:46:18 Coin[DEF]: Switch to DAG for epoch 544 09:46:18 Pool[DEF-0]: Received new job#: 19659f93 09:46:18 Krnl[00-DEF]: Create Kernel (A032#0-75) for NVIDIA GeForce GTX 1660 SUPER 09:46:18 Dag[00-DEF]: DAG(D544-32) is too large for this GPU. DAG size: 5.250 GB, available on GPU: 0 B. 09:46:18 Miner: **** No active GPU left for mining 'ETHASH'. TT-Miner terminates now.

What is the minimum memory requirements for each algo? Im this case GTX1660Super is a 6GB card, but the miner reports 0B.

[TrailingStop]

Hi,

thanks for the bug report. Do you have a screen attached to that GPU? When TT starts, it prints a status of the GPU before any memory is allocated. That looks like this: 09:29:17 #0 CudaId-2 PCI 01:00.0 nVidia NVIDIA GeForce GTX 1050 Ti 4.000 GB / 1.754 GB SM: 6 Compute: 6.1 09:29:17 #1 CudaId-0 PCI 02:00.0 nVidia NVIDIA GeForce RTX 3060 12.00 GB / 11.84 GB SM: 28 Compute: 8.6 09:29:17 * #2 CudaId-1 PCI 06:00.0 nVidia NVIDIA GeForce GTX 1070 8.000 GB / 7.892 GB SM: 15 Compute: 6.1

It shows build in memory and the memory that is available for TT. Can you please send a screenshot from this information? Thanks a lot.

[UselessGuru]

20:00:52 GeForce Driver-Ver: 527.56 20:00:52 Installed CUDA-Ver: 12.0 20:00:52 Supported CUDA-Ver: 11.8 20:00:52 Active CUDA-Ver: 11.8 20:00:52 20:00:52 CUDA devices: 3 20:00:52 * #0 CudaId-0 PCI 11:00.0 nVidia NVIDIA GeForce GTX 1660 SUPER 6.000 GB / 5.840 GB SM: 22 Compute: 7.5 20:00:52 #1 CudaId-1 PCI 12:00.0 nVidia NVIDIA GeForce GT 1030 2.000 GB / 922.7 MB SM: 3 Compute: 6.1 20:00:52 #2 CudaId-2 PCI 16:00.0 nVidia NVIDIA GeForce GTX 750 Ti 2.000 GB / 1.426 GB SM: 5 Compute: 5.0

[UselessGuru]

Version 2023.1.6:

23:23:50.468 CUDA devices: 3 23:23:50.473 * #0 CudaId-0 PCI 11:00.0 NVIDIA GeForce GTX 1660 SUPER 6.000 GB / 5.840 GB SM: 22 Compute: 7.5 23:23:50.473 #1 CudaId-1 PCI 12:00.0 NVIDIA GeForce GT 1030 2.000 GB / 910.2 MB SM: 3 Compute: 6.1 23:23:50.473 #2 CudaId-2 PCI 16:00.0 NVIDIA GeForce GTX 750 Ti 2.000 GB / 1.959 GB SM: 5 Compute: 5.0

23:23:50.663 Pool[FIRO-0]: Connecting to: stratum+tcp://firopow.mine.zergpool.com:3001 23:23:50.720 Pool[FIRO-0]: Connected to: firopow.mine.zergpool.com:3001 - IP4/IP6: 103.249.70.7 23:23:50.776 Pool[FIRO-0]: Set pool extra nonce to: 0x84B994 23:23:50.924 Pool[FIRO-0]: Difficulty changed from 0 H to 42.95 MH 23:23:50.929 Coin[FIRO]: Switch to DAG for epoch 481 23:23:50.929 Pool[FIRO-0]: Received new job#: 2200c 23:23:50.929 Pool[FIRO-0]: Working on Block#: 626093 - Blocks until epoch change: 507 23:23:50.929 Coin[FIRO]: Modify algo for block: 626093, ProgPoW period: 626093 23:23:50.929 Krnl[00-FIRO]: Create Kernel (A049#1E922-75) for NVIDIA GeForce GTX 1660 SUPER 23:23:50.099 Dag[00-FIRO]: DAG(D481-49) is too large for this GPU. DAG size: 5.258 GB, available on GPU: 0 B. 23:23:50.099 Miner: **** No active GPU left for mining 'FIRO'. TT-Miner terminates now.

DAG for Firo is only 4.77 GB

[UselessGuru]

Still an issue with 2023.1.7

[UselessGuru]

Any news?

[TrailingStop]

Should be fixed with the next version. Working on it.

[UselessGuru]

Any news?

[UselessGuru]

Any news?

[UselessGuru]

Any news?

[TrailingStop]

Hi - sorry for my late reply. I'm still working on a new version. I think this bug is fixed.

[UselessGuru]

I'm still working on a new version. I think this bug is fixed.

Any idea when you release a new version?

[TrailingStop]

Hi - I think this bug is fixed in the new beta - 2023.2.0. At the moment windows only, Linux will follow soon.