Gigahorse CPU Farming Issue

Chia Farming & Harvesting
1

Any help is greatly appreciated!

I am replotting everything after my previous wallet was compromised. I created a new wallet and a new plotnft. I have been plotting using the gigahorse cpu plotter using c9 compression. The plotting seems to be going ok, they are showing up under the plots, harvest, and pooling tabs in the gigahorse GUI. However, I am not getting any pool points and my pool (XCHPOOL) dashboard is not showing any activity. My debug log has a bunch of these errors:

2024-05-01T05:43:49.822 harvester chia.harvester.harvester: ERROR    Exception fetching qualities for F:\plot-k32-c9-2024-04-29-15-27-60c6219de375235564cb971ac709a37406df5db1a00c3ea5c6a331a4bd500ab1.plot: recompute failed to produce a candidate (table = 2) PlotInfo(prover=<chiapos.DiskProver object at 0x000002B172807770>, pool_public_key=None, pool_contract_puzzle_hash=<bytes32: 0438b205567545154ee9074f6615bdfc75d7b694bdff29123e5df502701c73ab>, plot_public_key=<G1Element 8b72176bb55b8df4b60d3c7a8200a30f0099206c0646c61d3ecd6a8fd52b1ca4dc3e30f08b24b52b019a34792b4b195e>, file_size=73197160990, time_modified=1714420862.2173378)
2024-05-01T05:47:08.849 harvester chia.harvester.harvester: ERROR    Exception fetching qualities for F:\plot-k32-c9-2024-04-29-20-53-b9ff17dd4981e08b22c99be48a884a89a8a212f0fc1ae521fcc76cd4a7a4cd3a.plot: recompute failed to produce a candidate (table = 2) PlotInfo(prover=<chiapos.DiskProver object at 0x000002B172804130>, pool_public_key=None, pool_contract_puzzle_hash=<bytes32: 0438b205567545154ee9074f6615bdfc75d7b694bdff29123e5df502701c73ab>, plot_public_key=<G1Element 8ff85dce2692da9357e245d2ba736e7c6bfaa25dbe5809360b1b0dfd48084a56007f0c7a46882140974dea957253c485>, file_size=73186864392, time_modified=1714440378.5892947)
2024-05-01T05:48:50.770 harvester chia.harvester.harvester: ERROR    Exception fetching qualities for F:\plot-k32-c9-2024-04-30-00-08-6958a51dbac5b23d1d846fa61f31c9b03f07404c843df2141570437453a17aef.plot: recompute failed to produce a candidate (table = 2) PlotInfo(prover=<chiapos.DiskProver object at 0x000002B1728067B0>, pool_public_key=None, pool_contract_puzzle_hash=<bytes32: 0438b205567545154ee9074f6615bdfc75d7b694bdff29123e5df502701c73ab>, plot_public_key=<G1Element b07b981b1db7b1d2ed9f7e692677887308f7354e38095b67d729f0456d1c24eb0c14a96c16ac2dcd0a96f362b14ca243>, file_size=73196839104, time_modified=1714452067.2099657)
2

Here are two possibilities:

As it pertains to:

I hope that your box is not compromised.
Did you re-install your OS, or take some other action to ensure that your box is clean?

3

Where do I enter CHIAPOS_MAX_GPU_DEVICES=0 ?

Yes, I reformatted the system drive and reinstalled windows and everything else.

4

after running proofofspace lookup, i’m getting the following error: “Threw: recompute for this level not supported on given hardware”

5

In the environment variables.

6

I added it to environmental variables, and when i run proofofspace lookup, it says it is using 16/16 CPU threads, 0/0 CUDA devices, and 0/2 OpenCL devices (AMD Accelerated Parallel Processing)

7

Does gigahorse actually support C9 using CPU farming? Has anyone confirmed this?

8

Have you checked that table - GitHub - madMAx43v3r/chia-gigahorse.

Most likely, you can add nVidia k2200 ($20-30) that will take care about those CPU compatible compressions using most likely less power than your CPU.

9

The CPU plotter says it can plot up to C9, I assume you’d also be able to farm these plots using the same hardware, no?

10

If you go back to that table, it looks like c8/9 may not be CPU farmable. On the other hand, c11-13 are. Again, the reason you see max farm sizes dropping (third column from the last) is that CPU cycles go up with every c-level. On the other hand, k2200 is more energy efficient for most CPUs, I think (not sure about the latest), especially toward the higher end of those ranges.

11

Thanks, I decided to sell my 7800xt and buy a 4070. I’ll have it by tomorrow and can start plotting C31 with that. Can I do C32 or C33 with the 4070?

12

Well, that farming table is your best info (if you search the forum, there is also a google doc where people provide their results). You should be able to compare 4070 performance to 4070S that is listed there to get an estimate of the max farm size you can have with those formats and where your farm fits.

13

As long as your 4070 is 12GB then you can farm C32, C33 requires 24G of VRam, its in the table linked to below.

You should also be aware of just how much you can farm, for example a 4070S can only farm around 200 TiB of raw capacity of c32 plots after the filter change next month.

14

I currently only have 70-80 TB of raw capacity. Do you think I could fill it all with C32 and be ok after next month?

15

With the 4070, easily.

1 Like