Claymore's CryptoNote AMD GPU Miner v11.3

[pehoko]

On my 380 4G claymore 10.2 works but sometimes stops accepting shares after few hours. So I am with sgminer - no problems but speed is slower.

[Thekingace]

Hi,

I tried switching to monero from eth mining and it works fine on one of my rig getting 800h/s per gpu but my other rig gets 300ish mhs on all card for 1 second than all but the first card drop 2 0mhs. They're both identical except for GPU brands XFX GTS 580s 8gb vs nitro+ 580 8gb.

Any idea? Ethereum is working fine on both.

[fgarpe]

Tha latest version run perfectly under Linux with 2 RX580 8Gb

https://i.imgur.com/xzYzwcc.png

Is there any plan to add dual mining support?

[Call_Me_Bambi]

Hi,

I tried switching to monero from eth mining and it works fine on one of my rig getting 800h/s per gpu but my other rig gets 300ish mhs on all card for 1 second than all but the first card drop 2 0mhs. They're both identical except for GPU brands XFX GTS 580s 8gb vs nitro+ 580 8gb.

Any idea? Ethereum is working fine on both.

Check your virtual memory is set high enough, from what I can remember Claymore's Cryptonote miner needs 8GB virtual memory PER CARD to work , or at least work to its best abilities.

Try your setup with a different miner, this might help fault find where the problem is. I prefer to use SGminer for Monero https://github.com/genesismining/sgminer-gm/releases .

[noobminer001]

Tha latest version run perfectly under Linux with 2 RX580 8Gb

https://i.imgur.com/xzYzwcc.png

Is there any plan to add dual mining support?

Hello,

What are your settings for the RX580 8G? Core Speed/Mem Speed/Voltage? What do you have your -h set at? Are you running SMOS?

I'm wondering if you are able to get it stable for longer than 24hrs. I can usually get my RX580 8G rigs stable for 5-8 hrs but then it will reset so I end up getting 3-4 resets in a day. When I was mining ETH with the same rig, it would be stable for 24hr+. Right now my settings are 1200/2000/1000mV with a -h of 1600 running SMOS. I am also running RX570 4G rigs and they are a little more stable at 1-2 resets a day. Not sure why the higher memory cards are having issues. Settings for the 4G cards are 1200/1750/1000mV with a -h of 850 running SMOS.

Both of my rigs are using 11 GPUs, 8G of RAM, z270 chipset.

Cheers,

[leemaster]

It would be very good to have a nvidia version

[invicta]

Tha latest version run perfectly under Linux with 2 RX580 8Gb



Is there any plan to add dual mining support?

Hello,

What are your settings for the RX580 8G? Core Speed/Mem Speed/Voltage? What do you have your -h set at? Are you running SMOS?

I'm wondering if you are able to get it stable for longer than 24hrs. I can usually get my RX580 8G rigs stable for 5-8 hrs but then it will reset so I end up getting 3-4 resets in a day. When I was mining ETH with the same rig, it would be stable for 24hr+. Right now my settings are 1200/2000/1000mV with a -h of 1600 running SMOS. I am also running RX570 4G rigs and they are a little more stable at 1-2 resets a day. Not sure why the higher memory cards are having issues. Settings for the 4G cards are 1200/1750/1000mV with a -h of 850 running SMOS.

Both of my rigs are using 11 GPUs, 8G of RAM, z270 chipset.

Cheers,

I would hazard a guess that your memory is too far OC'ed. Cards that were stable on ETH for me, have been less tolerant under Cryptonote or Equihash - maybe because both of those are much more memory hard. To be honest, this exact reason is why I stopped using Nicehash Miner and just configure the software to point to their servers as I found when it switches settings, one set of OC/voltage/config for cards that worked in ETH might crash them when it flips to another algo.

[noobminer001]

Tha latest version run perfectly under Linux with 2 RX580 8Gb

https://i.imgur.com/xzYzwcc.png

Is there any plan to add dual mining support?

Hello,

What are your settings for the RX580 8G? Core Speed/Mem Speed/Voltage? What do you have your -h set at? Are you running SMOS?

I'm wondering if you are able to get it stable for longer than 24hrs. I can usually get my RX580 8G rigs stable for 5-8 hrs but then it will reset so I end up getting 3-4 resets in a day. When I was mining ETH with the same rig, it would be stable for 24hr+. Right now my settings are 1200/2000/1000mV with a -h of 1600 running SMOS. I am also running RX570 4G rigs and they are a little more stable at 1-2 resets a day. Not sure why the higher memory cards are having issues. Settings for the 4G cards are 1200/1750/1000mV with a -h of 850 running SMOS.

Both of my rigs are using 11 GPUs, 8G of RAM, z270 chipset.

Cheers,

I would hazard a guess that your memory is too far OC'ed. Cards that were stable on ETH for me, have been less tolerant under Cryptonote or Equihash - maybe because both of those are much more memory hard. To be honest, this exact reason is why I stopped using Nicehash Miner and just configure the software to point to their servers as I found when it switches settings, one set of OC/voltage/config for cards that worked in ETH might crash them when it flips to another algo.

I had the same issue and I think I discovered why - Cryptonight uses all of the GPUs memory while ETH only uses ~2.3GB. This means the memory needs more power on cryptonight.

Long story short: increase the power limit on the GPUs. I had mine at 0 for ETH but +5-10 makes my cards stable for Cryptonight.

Interesting, I didn't know that ETH only uses 2.3Gb of memory.

As for the OC'd cards, 2000mHz is actually stock speed so unless I need to underclock to solve this issue.

When you've increased the power limit, is that the memory voltage or just the overall power level? In SMOS I have two options which is why I'm asking. Do you know what voltage level you were running the cards? I'm at 1000mV right now, which is undervolted I believe. I'm not sure what stock voltage levels are (maybe 1300 or 1400?)

Thanks!

[invicta]

Tha latest version run perfectly under Linux with 2 RX580 8Gb



Is there any plan to add dual mining support?

Hello,

What are your settings for the RX580 8G? Core Speed/Mem Speed/Voltage? What do you have your -h set at? Are you running SMOS?

I'm wondering if you are able to get it stable for longer than 24hrs. I can usually get my RX580 8G rigs stable for 5-8 hrs but then it will reset so I end up getting 3-4 resets in a day. When I was mining ETH with the same rig, it would be stable for 24hr+. Right now my settings are 1200/2000/1000mV with a -h of 1600 running SMOS. I am also running RX570 4G rigs and they are a little more stable at 1-2 resets a day. Not sure why the higher memory cards are having issues. Settings for the 4G cards are 1200/1750/1000mV with a -h of 850 running SMOS.

Both of my rigs are using 11 GPUs, 8G of RAM, z270 chipset.

Cheers,

I would hazard a guess that your memory is too far OC'ed. Cards that were stable on ETH for me, have been less tolerant under Cryptonote or Equihash - maybe because both of those are much more memory hard. To be honest, this exact reason is why I stopped using Nicehash Miner and just configure the software to point to their servers as I found when it switches settings, one set of OC/voltage/config for cards that worked in ETH might crash them when it flips to another algo.

I had the same issue and I think I discovered why - Cryptonight uses all of the GPUs memory while ETH only uses ~2.3GB. This means the memory needs more power on cryptonight.

Long story short: increase the power limit on the GPUs. I had mine at 0 for ETH but +5-10 makes my cards stable for Cryptonight.

Interesting, I didn't know that ETH only uses 2.3Gb of memory.

As for the OC'd cards, 2000mHz is actually stock speed so unless I need to underclock to solve this issue.

When you've increased the power limit, is that the memory voltage or just the overall power level? In SMOS I have two options which is why I'm asking. Do you know what voltage level you were running the cards? I'm at 1000mV right now, which is undervolted I believe. I'm not sure what stock voltage levels are (maybe 1300 or 1400?)

Thanks!

2.3GB is the current size of the DAG file (epoch 150 I think). Cryptonote definitely uses more (it tells you on initialisation exactly how much) and hammers the memory controller a lot more, it's a lot less tolerant of overclocking too. 1050-1100mV seems to be 'normal' voltage for the R9 and RX cards when in 3D mode.

[swalayan]

Thx for share, my rx470 4g get 820 with cryptonight 9.7. New version get 830 but not stable, all with same setting. I dont know why. My mem volt 1080, gpu temp 60, do u think its still safe for long term use? Thx u, love u all..

[Ursul0]

some bad juju on the 370 in slot with display dummy.
Got 4 more of 470 on the same platform - doing just fine, but this one produces rejects 1/4 of the time.

[tg88]

Could we get an update to 9.7 which allows for mining of Electroneum? I've got older cards too. 10.2 just doesn't seem to work well with the older stuff.

You can test the -dmem 0 option in 10.2 if you are experiencing instability.

or if you want, it is also possible to mine electroneum in version 9.7 without problems.

[Ursul0]

some bad juju on the 370 in slot with display dummy.
Got 4 more of 470 on the same platform - doing just fine, but this one produces rejects 1/4 of the time.

Also have issues with starting and never stopping to prodiuce "low difficulty share" on vegas
so... fuck it!
370&470 back to sgminer - better rate and almost 0 rejects
and vegas are doing well on stak (same rate, lower dev fee)

EDIT: the reason to use claymore was the convenience of monitoring (EthMan) from a single point of access - not worth it in this case

[invicta]

some bad juju on the 370 in slot with display dummy.
Got 4 more of 470 on the same platform - doing just fine, but this one produces rejects 1/4 of the time.

This has been done to death by me and many others.

Pitcairn core (HD7850, R9 270/X, 370/X) is broken on Claymore 10.2 Beta. Run 9.7 or remove that Pitcairn card from your rig and run 10.2 on the Ellesmere RX cards only. If I were in your position I would do the latter as 10.2 is faster, cooler and uses less power than 9.7.

I would use SGminer or STAK but don't have the patience as I've heard it needs a lot of setting up.

[lpedretti]

GPU 3, GpuMiner dk2a failed -4

3 of my 6 gpus rig always throw me this error, but the specific gpu changes with each restart

Any clues?

[lpedretti]

GPU 3, GpuMiner dk2a failed -4

3 of my 6 gpus rig always throw me this error, but the specific gpu changes with each restart

Any clues?

Found the problem, i needed to raise windows virtual memory

[Ursul0]

some bad juju on the 370 in slot with display dummy.
Got 4 more of 470 on the same platform - doing just fine, but this one produces rejects 1/4 of the time.

This has been done to death by me and many others.

Pitcairn core (HD7850, R9 270/X, 370/X) is broken on Claymore 10.2 Beta. Run 9.7 or remove that Pitcairn card from your rig and run 10.2 on the Ellesmere RX cards only. If I were in your position I would do the latter as 10.2 is faster, cooler and uses less power than 9.7.

I would use SGminer or STAK but don't have the patience as I've heard it needs a lot of setting up.

Yep, back to running SGminer. no issues for the last 30+ hours: total 3 hw errors and below 3% rejects... with higher hash and 0 devfee:)

[jblewis85]

Hello,

Im mining Some electronium but i am getting Low hashrate on my rx 580 8gb cards. My 570 4gb and 580 4gb cards get ~700m/h. The 8gb cards a get ~350 mh? Not sure why this is. All card have bios mod for ether mining. I set my vram to 50000. I was trying to clock the cards one by one with claymore but it does not seem to take the mclock a ccclock. Cryptonote seems much laggy and slower to start mining on all cards then ether as well.

start.bat below nothing in config.
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_SINGLE_ALLOC_PERCENT 100
setx GPU_MAX_ALLOC_PERCENT 100

NsGpuCNMiner.exe -xpool electroneum.hashparty.io:5555 -xwal etnkGQeNyQ3aJ3iypn7mR1iu4HbTqpfQvMGQvYsVUHnG3pRxEv5xyyJ7jvErPYjFNBVQXah6eae8fdq ckYYExGwH7Bw9XMHkng -xpsw x  -nofee 0 -ftime 10 -h 0 -dmem -1 -cclock 1150,1380,1150,1366,1366,1150 -mclock 1900,2150,2050,2000,2000,1950

 any thoughts my guess is its something to do with the mem clock being higher on the 8gb cards but it wont accept mods. thanks

[invicta]

some bad juju on the 370 in slot with display dummy.
Got 4 more of 470 on the same platform - doing just fine, but this one produces rejects 1/4 of the time.

This has been done to death by me and many others.

Pitcairn core (HD7850, R9 270/X, 370/X) is broken on Claymore 10.2 Beta. Run 9.7 or remove that Pitcairn card from your rig and run 10.2 on the Ellesmere RX cards only. If I were in your position I would do the latter as 10.2 is faster, cooler and uses less power than 9.7.

I would use SGminer or STAK but don't have the patience as I've heard it needs a lot of setting up.

Yep, back to running SGminer. no issues for the last 30+ hours: total 3 hw errors and below 3% rejects... with higher hash and 0 devfee:)

Do you have the download link and settings for it? I used Claymore as it's convenient and I know it, but I thought SGminer wasn't in development any more, or does that not matter now?

[invicta]

Hello,

Im mining Some electronium but i am getting Low hashrate on my rx 580 8gb cards. My 570 4gb and 580 4gb cards get ~700m/h. The 8gb cards a get ~350 mh? Not sure why this is. All card have bios mod for ether mining. I set my vram to 50000. I was trying to clock the cards one by one with claymore but it does not seem to take the mclock a ccclock. Cryptonote seems much laggy and slower to start mining on all cards then ether as well.

start.bat below nothing in config.
setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_SINGLE_ALLOC_PERCENT 100
setx GPU_MAX_ALLOC_PERCENT 100

NsGpuCNMiner.exe -xpool electroneum.hashparty.io:5555 -xwal etnkGQeNyQ3aJ3iypn7mR1iu4HbTqpfQvMGQvYsVUHnG3pRxEv5xyyJ7jvErPYjFNBVQXah6eae8fdq ckYYExGwH7Bw9XMHkng -xpsw x  -nofee 0 -ftime 10 -h 0 -dmem -1 -cclock 1150,1380,1150,1366,1366,1150 -mclock 1900,2150,2050,2000,2000,1950

 any thoughts my guess is its something to do with the mem clock being higher on the 8gb cards but it wont accept mods. thanks

My cards did this the other day. It was Windows 10, ever thinking it's being helpful, it updated my Beta Blackchain driver to Crimson ReLive 17.7.2 (or similar) and didn't tell me, then yes, my cards became very laggy, slow to initialise and the RX 580 went from 600+ down to 315H/s. I would look there first.