tdxminer has become teamredminer

Please go to the teamredminer thread for new releases.

Thanks.  It's nice to hear people are enjoying the miner

Adding support for rx460/560 with the amdgpu-pro drivers would take some effort, and I'm currently focused on another project.
However, I *think* they should work with the miner as is on the ROCm drivers.  The ROCm 1.8 drivers remove the requirement of PCIe link width for Polaris cards, so typical x1 risers work, but they will need to run at PCIe gen3.
I have not tested any rx460/560 as I don't have hardware available.

Right, tdxminer is not available by default in HiveOS and you will have to download it.
If you're using a graphical interface, the distro should have a browser and archive extractor installed.
If you're using a terminal interface, you can run the following command to download and extract the archive:
In either case, you will need a terminal to run the miner.

Does your machine have an internet connection?


Ubuntu will not work right out of the box since it won't have drivers.  HiveOS will have amdgpu-pro drivers for Polaris GPUs out of the box.  For Vega you're best off installing Ubuntu and then installing drivers from the ROCm 1.7 repos.

Polaris with amdgpu-pro drivers will work with gen2 pcie.  Vega on ROCm will not work on gen2 pcie (requires gen3).

The dev pool failing to connect is my fault.  There was an error that took out the dev pool for a few hours, it is back up now.  You may occasionally see these errors from time to time as I need to restart it to push updates, but I do my best to keep the down time to only a few seconds.

I don't have much else to suggest for stability.  Just make sure your temps are good, and if you're not overclocking/undervolting the cards should be stable.

This is most likely due to the card crashing.  This can happen due to unstable overclocks and/or underclocks.  This can also happen if the GPU goes over temp.  I suggest monitoring the GPU's power state and temperature.  With the ROCm drivers installed, I find the rocm-smi tool to be the easiest way to do this.


Sorry crocozino, but I'm not actively working on the Windows version right now.  All the time I've been able to put into the miner have go to fixing bugs lately.  I will not be able to get a windows version out any time soon.  I suggest trying one of the many linux mining distros and reading/following one of the many linux mining guides out there.

Ugh, government IP blocking is such bs.

Sorry fellas.
I will not be setting up additional dev servers at this time.
You'll either have to run with the reduced hash rate, or find a way to get around the IP block.

Can you elaborate on what you mean by not available?  Does DNS fail to resolve the domain?  Is the route blocked by your ISP?

What do you get if you try to ping dev.tdxminer.com?  Maybe run a traceroute too.

The message you are seeing is because the miner cannot connect to the dev pool.
The miner connects to the dev pool via port 443.  Please check that your firewall allows this port.  (This is usually the https port for connecting to websites with encryption)

This is the bug that never dies.

New release posted with more bug fixes and better handling of pools that stop responding.
Please let me know if this version still gives you any problems.

I'll try to put together something easier to run later tonight.

You appear to have skipped the steps of setting up the ROCm apt repository, and the part about installing rocm-opencl/rocm-opencl-dev.  Possibly also the part about adding your user to the video group.
You can't just skip to the end of the instructions and expect it to work.

The amdgpu-pro and ROCm are entire separate driver stacks.  You should install one or the other.  Installing both will most likely lead to problems.

I suggest you uninstall ROCm and amdgpu-pro and reinstall amdgpu-pro using the command:
./amdgpu-install --opencl=legacy
as per the instructions on AMD's website: https://support.amd.com/en-us/kb-articles/Pages/Installation-Instructions-for-amdgpu-Graphics-Stacks.aspx

After this, clinfo should show your 580s as devices named "Ellesmere"

I guess I assumed users would be familiar with linux.  The ROCm page that is linked in the main post should have instructions for how to install it.
I can try to find some relevant linux mining guides and post additional steps to those.  I'm not sure how much time I have to do all that though.


Not good.  I haven't made as much progress as I'd like and some things have come up recently that are making it hard to make progress.
Lately fixing the few bugs reported is all the work I've been able to put into the miner.


What does clinfo have to say?

A new release is up with bug fixes for the recently posted network errors.

As before,
I advise anyone using version 0.2.1.1 or earlier to monitor their rigs closely in case of this error occurring.  If your pool supports alert/monitoring emails, please use them.

I just saw this error on one of my miners earlier today as well.  I am investigating it.

I will be doing some maintenance to the miner's dev pool tonight.
You may see the miner reporting that it cannot connect to the dev pool.
This should only last a few minutes.

If your miner continues to report problems connecting to the dev pool, please restart it and update to v0.2.1.1 when you can, as this version has some network bug fixes.

My experience with MPH has not been great.  In the process of developing the miner I've found two bugs in their XZC pool.  One of which allows a user exploiting the bug to get credit for a higher share rate than they are actually submitting, while at the same time not contribute to actually finding a block.  I've reported the bugs to the MPH guys, but they have yet to get back to me about the bug that basically allows cheating.

I've used suprnova's pools before, and have had a great experience.  However, their XZC pool seems to have some issues that make me worry: the stats on the pool seem to be completely broken, the hash rates it reports are ~30% higher than actual hashrates, it reports previous block effort incorrectly making the pool look like it has better historical returns than it actually has (any block that took more than 4,294,967,295 shares is reported as taking 4,294,967,295 shares).

Ominous network is new and they seem to currently have great support and 0% fee as they are trying to attract miners.  I ran a rig against them for a week and had no problems, also no returns since I was the only one on the pool.  The way the pool issues jobs can result in some rejected shares, due to every new job invalidating all previous jobs.  But in my time running this rejection rate was low, under 1%.  This will be more of an issue for miners running with high network latencies to the pool.

I had tried to run with f2pool, but the language barrier slowed me down a bit.  Also I couldn't find a web front for their pool stats, which is something I would like to have in the pools I use.

If enough of you guys can agree on a good pool to use, I would be happy to redirect my rigs and the dev fee hashes to it to help total hashrate.  Personally I'm leaning towards either Ominous or Mintpond, what do you guys think?