So anyone able to Run PhoenixMiner on Windows 10 and RX570/470 with 4GB?

Happy for you. Could you please post your config settings? As you can see from my post I get 0 MH/s for my 4 x RX570 4G and I had to switch to lolminer which is unstable:

Hello.

I'm using PhoenixMiner 5.2c on Windows 10 with Adrenaline 19.7.5 drivers,  4 x 570RX 4GB cards in "compute mode" (VGA output connected to intergrated Intel GPU) with following settings:
Yet after DAG generation I get 0 Mh/s on every card, please advise what could be wrong, thank you.

Is there an option for miner restart when hashrate drops and/or one or more of the cards in the rig hashrate drops to 0? I've encountered this issue for a few times now and this didn't happen with other miners :/

Can you please tell me which version of old Crimson beta driver are you using now? Will 17.11.4 do?

I too face 0MH/s with my RX570/470 4G since today on Windows 10, Adrenaline driver v19.20. Only monitor is connected to integrated mobo graphics (HDMI port), yet Windows still not allowing to allocate more than 3737 MB of VRAM (according to HWInfo). -eres 1 or 0 doesn't help, lowering GPUs' memory timings to 1700 (instead of ~2000) gives no change.

From @nanopool_org twitter:

Does anyone experience random (in time periods) total system freezes on Claymore 10.3 on Win 10 x64 (Aug 23 blockchain drivers with 3 x RX570 and 1 RX470)?
So only reboot via mobo helps.

So far the most I got it running without freezes is 48 hours, but it can freeze in 2 hours or 10 or x hours.
There's nothing in Claymore or Windows logs.

I've experienced exactly the same behavoir when mining XMR with Claymore's Cryptonote AMD Miner v11 - but not on v9 or v10 - they were quite stable.
My cards are BIOS moded, overclocked in memory, downclocked in core and core voltage - last few days I'm bringing each card 20 MHz down after another freeze trying to understand if it could be related to memory - so far got to 1960 MHz.
However, I think if it was to do with memory - I'd get memory errors (which I don't) and/or general GPU drivers failure or single/multiple GPUs hashing stops - but it just freezes whole system. Quite frustrating tbh.

Btw I have 32G of page file, so virtual memory shouldn't be the cause.
I could try Claymore dual miner v10.5, but something tells me the freezes won't go away :/

Yeah, I know -wd 1 is default, just had to mention it's 100% enabled.

Actually I had "quickedit mode" enabled, now I've disabled it. When you're saying "WD used to hang" - do you mean it worked, but the miner was hanging/freezing when being restarted by WD or WD wasn't performing any kind of action until button was pressed (like in my case)? Thx for the tip though.

Using Claymore 10.3 on Win 10 x64, Aug 23 blockchain drivers with 3 x RX570 and 1 RX470. The problem I'm facing is miner's watchdog doesnt seem to be working - when one (or multiple) GPU gets "GPUx hangs in OpenCL call" error - the miner just stops working - until I hit "s" on keyboard, then and only then it restarts.

This is quite frustrating. Watchdog is enabled (-wd 1) - it even says so when miner launches, also "-r 15" is set. Any idea/solution to this? Thanks.

Hi. I'm on win 10, 64 bit, using SGMiner 5.5.5.
Currently I have 2 GPUs: Sapphire RX 470 Nitro+ and Gigabyte RX 570 Aorus.
I'm experiencing the strangest thing: when one of the cards is connected to PCIe x16 slot (number 1) - be it using riser or directly - it provides bigger and more stable hashrate.

Example:
- RX 470 is in PCIe x16 slot (via riser) - gives ~900 H/s stable, no slowdowns after 20+ hours;
- RX 570 is in PCIe x1 slot (via riser, the one just above PCIe x16), - gives ~860 H/s on average, but with slowly degrading performance. It could go to 840 or even less gradually.

Now I know one might think "you have different overclock settings, BIOSes, the GPUs are different after all", BUT! When I switch PCIe connections of these cards - RX 570 starts to mine w 900 H/s stable, and RX 470 drops as described.
So could PCIe x16 be the difference here? I don't see how though, since both of the cards are connected via PCIe x1 risers. Thanks.

EDIT: there's one more thing - none of the included OC settings are working, for example "gpu-vddc" doesn't set (I set it in global part of the config). I believe it's to do with the drivers, right? I use latest crimson blockchain for W10.
However, there's another issue - SGMiner resets core voltage of all GPUs every so often. This happens with MSI Afterburner closed and/or open. I don't face such an issue with any of claymore's miners. Anyone has a solution for this? Thx.

Consider yourself lucky. This version keeps crashing my rig so it needs hard reset every 24 hours or even more frequently. I'm getting rid of it for now. Not to mention Claymore "realised" there was a "bug" that resulted in dropping old jobs with delay:


It's been almost 5 days and no update from Claymore.

I've been using claymore v11 for around one day, and yes it seems there's hashrate increase - judging from the miner info.
However, I've been keeping an eye on actual statistics of the pool I'm mining on - and it looks like hashrate is actually quite lower.

To give an example - 2 of my GPUs (RX 570 btw) are mining with claymore v9.7 with 1650 H/s total on average (stat from the miner) - pool statistics are similar.
Another 2 GPUs are mining with claymore v11 with 1830 H/s (same pool, BIOS, OC settings).
However, if I bluntly compare average shares per hour ratio (I know the difficulties can differ though), I get around 375 for BOTH sets of GPUs on v9.7 and v11.

I'm starting to get suspicious (and a bit paranoid) about v11.

Thanks, "-r" option did the job it seems: I've set it to 15 minutes and after another single rejected share the miner proceeded to reconnect without my input.

Hi! I'm using Claymore's GPU Miner v9.7, running it on Win 7, 64bit.
When I get my share rejected (which happens on average once in a 24 hours) by the miner, it seems it doesn't attemp any reconnection. I have to press "s" in the command line window and only then it reconnects.

How can I make it do that automatically? I've read the "Readme!!!.txt" and miner's github docs, but there's nothing about it. Thanks.