Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux)

[leonix007]

Guys, dont flame me here, but I just added some AMD cards to me Nidvia rigs. Is it possible to use claymore for just the AMD cards and another miner for nidvia?

Yes, check GPU indexes numbers on AMD and use -di option, been using this on one of my rigs, Claymore on AMD, EWBF on nvidia's

[1714111155]

1714111155

[1714111155]

1714111155

[1714111155]

1714111155

[Popaj]

Hi I am new here and I have problem.
I recently tested Claymore: Version 11.1 and new one 11.3. In the first 12-14 hours it shows perfect performance and stability. After that, it extinguished and immediately after each attempt to start, it extinguishes.
After starting again bat. file it just close and that version can not start again.


http://prntscr.com/ipo245
 I have 6x Sapphire rx 580 Nitro+
On Versio 11.0 work stabile.

[BTC22]

Guys, dont flame me here, but I just added some AMD cards to me Nidvia rigs. Is it possible to use claymore for just the AMD cards and another miner for nidvia?

Yes, check GPU indexes numbers on AMD and use -di option, been using this on one of my rigs, Claymore on AMD, EWBF on nvidia's

Can I PM you?

[whoismoses]

Anyone having issues with 11.3 and Hawaii GPUs? Every 10-15 minutes my rigs will lose 5-7% speed for about 2-3 minutes, and then they go back to normal. This happens continually, on all my Hawaii based rigs. Never saw this issue with 10.6 that I just upgraded from.

Check to see if they are getting thermal throttled.

[Zerb]

Hi I am new here and I have problem.
I recently tested Claymore: Version 11.1 and new one 11.3. In the first 12-14 hours it shows perfect performance and stability. After that, it extinguished and immediately after each attempt to start, it extinguishes.
After starting again bat. file it just close and that version can not start again.


http://prntscr.com/ipo245
 I have 6x Sapphire rx 580 Nitro+
On Versio 11.0 work stabile.

I have the same problem, someone can help us

[Ursul0]

The current AMD drivers in compute mode are optimized for Windows 10 v1709 FCU. Those are the fastest drivers to use with RX 570/580 cards. The only reason to use the blockchain drivers on Windows v1709 with RX 570/580's is some old miners like SGminer don't work with the latest drivers.

In my experience, in what mining is concerned, the Beta Blockchain drivers are identical to all new drivers. No speed difference whatsoever.
If someone has proof otherwise please come forward.

I have a six card RX 570/580 rig with dual boot Windows 10 v1709 partitions. The latest Adrenaline drivers on one and the Aug 23rd blockchain drivers on the other, so I can run older miners that are not compatible with the current drivers if I need to. Using the same settings for both, in addition to the blockchain drivers being noticeably slower by 3-4 MH/s, they have frequent drops in hash rate and issues with Windows as well. The AMD blockchain drivers came out before the Windows 10 v1709 FCU and are NOT supported. The v1709 FCU was a major update to how Windows interacts with GPU's and also expanding the previous GPU limit from 8 to 13, which is why AMD had to release new drivers that were compatible with the FCU.

The bold part is factually incorrect. As I have multiple miners on anniversary edition (ver 1607 build 14393.0 )
The key to 12 GPUs is the driver version, if I remember correctly it's about 17.10.1 when 8+ GPUs became supported.
As for the "blockchain drivers" - it doesn't support over 8 AMD GPUs, somewhat less stable and possibly has some issues with driver api  

[leonix007]

Guys, dont flame me here, but I just added some AMD cards to me Nidvia rigs. Is it possible to use claymore for just the AMD cards and another miner for nidvia?

Yes, check GPU indexes numbers on AMD and use -di option, been using this on one of my rigs, Claymore on AMD, EWBF on nvidia's

Can I PM you?

Go ahead

Hi I am new here and I have problem.
I recently tested Claymore: Version 11.1 and new one 11.3. In the first 12-14 hours it shows perfect performance and stability. After that, it extinguished and immediately after each attempt to start, it extinguishes.
After starting again bat. file it just close and that version can not start again.


http://prntscr.com/ipo245
 I have 6x Sapphire rx 580 Nitro+
On Versio 11.0 work stabile.

I have the same problem, someone can help us

If you are using windows 10, check if it has been updated, and probably windows defender or firewall blocked it. disable those two

[Viaggiatore]

Dear @Claymore,

I have noticed increase in stale shares from 11.2 to 11.3 about 3%. Nothing is changed beside the version of the miner. Can you check that?

What pool do you use?

us1.ethermine.org:14444, like before

I see some crazy values for stales today, up to 15%, it can be some issue on pool side. Try to start v11.2 and see if it helps.

Same for me but I'm using ethermine SSL server

[Viaggiatore]

@Claymore
did you notice a lower hashrate using SSL ethermine server? Tested 11.3 + SSL and noticed again (also 11.1 was acting in the same way) a lower hashrate than no SSL server.

Thanks

[Viaggiatore]

In 24 hours I'm getting 0 rejected eth shares and about 22 incorrect per card. 30 rejected decred per card. 0 memory errors. On 11.3. Is this acceptable?

yes

It is acceptable but 22 incorrect is about 4%. You would have a higher profit slowing the cards down and eliminating those errors. A 29mhs card with 0% errors has higher profit than a 30mhs card with 4% errors.

I'd start with lowering each mclock by just 20Mhz and see if that clears up the error rate.

Of my 17 cards I tolerate an incorrect shares rate of 1% on one card that I can't shake no matter what timings I use and 0.5% on 2 others. The rest can all go weeks without an incorrect share.

I never had incorrect ETH shares, all of my 8 cards have "0" memory errors. Now with 11.3 I'm getting ETH incorrects share, always at GPU 0 and GPU 4 (again with zero memory errors)...@Claymore something definitely wrong with this version

[rigomortis]

v11.3 is using SSL for devfees.  So if you are running a nodevfee you won't be getting those hashes.  Claymore will.  He's upped his game to get more shares from you!  Just masking it will higher hashrates for his newer version. 

[Bl4ckHorse]

You need to find the best settings for your own cards and memory type. My cards are a mix of Asus RX 580/480 with Samsung memory (first three with 2050 MHz memory overclock), Sapphire Nitro+ RX 580's with Samsung memory (2100 MHz memory overclock), Hynix memory (2200 MHz memory oveclock) and a MSI Gaming X RX 580 with Samsung memory (last card with a 2050 MHz memory overclock).

Code:

-cclock 1200,1150,1200,1200,1200,1200,1200,1200,1200,1200,1200,1200,1200
-mclock 2050,2050,2050,2200,2100,2100,2100,2220,2220,2220,2100,2220,2050
-powlim -10,-5,-10,-15,-10,-15,-10,-10,-5,-10,-10,-10,-10
-cvddc 950,950,950,850,900,850,850,900,900,900,850,900,900
-mvddc 950,950,950,850,900,850,850,900,900,900,850,900,900

Hello m8,
are these clocks and voltages for dual mining or solo ?
I have 6 of Sapphire rx 580 se with micron memory blue ones.
Nearly all of cards have up %75 asics quality. Whats the best values of clock mlock and dcc values for dual and solo eth mining ?
Im using -cvdcc 865,875,865,865,865,865 -cclock 1240,1240,1240,1240,1240,1240 -mclock 2250,2240,2245,2245,2250,2230 -mvdcc 875,875,880,875,880,880 and i don't use -powlim. Should i use powlim and these values is high or ok ? Hwinfo don't show errors. I also use -ethi 16. I want dual mining. Which values should i use both solo and dual mining ? I got 32.4 - 32.0 mh/s total is 193.4 mh/s. Hwinfo don't show errors but share is really so different. Sometimes calculated 350 mh/s sometimes 58 mh/s average on nanopool. Avarage values always under 190mh/s. I did 48 hours test ethermine and nano result same always 48 hours under 190 mh/s calculated.
Thanks for reply

[Metroid]

v11.3 is using SSL for devfees.  So if you are running a nodevfee you won't be getting those hashes.  Claymore will.  He's upped his game to get more shares from you!  Just masking it will higher hashrates for his newer version. 

Claymore did on v11.2, by the way phoenix miner has a lower devfee 0.65% and the tests i did showed phonexminer is legit on the devfee.

[cryptoillusion1]

What does this mean in the log:the lines in bold

00:47:50:494   860   ETH: 03/13/18-00:47:50 - New job from us1.ethermine.org:4444
00:47:50:494   860   target: 0x0000000112e0be82 (diff: 4000MH), epoch 174(2.36GB)
00:47:50:494   860   gpu #0 dt 0.13 (0%, good)
00:47:50:494   860   gpu #1 dt 0.13 (0%, good)

[meru890]

Dear @Claymore,

I have noticed increase in stale shares from 11.2 to 11.3 about 3%. Nothing is changed beside the version of the miner. Can you check that?

What pool do you use?

us1.ethermine.org:14444, like before

I see some crazy values for stales today, up to 15%, it can be some issue on pool side. Try to start v11.2 and see if it helps.

Same for me but I'm using ethermine SSL server

same here, but running 11.0 with eu1:ethermine.org4444 and today got around 10% stale shares, didn't change any settings at all

[Claymore]

v11.4:

- fixed issue with stale shares.
- fixed issue with incorrect shares.
- a few minor bug fixes and improvements.

This is highly recommended update.

[skablast]

Just finished the test at the WALL

I am sticking with the previous version.

V11.3 does offer a little increase in performance but it come with a Energy demand increase that offset the increase in hashrate.
Overall, I was earning more/card/day a week ago

[big al]

Just finished the test at the WALL

I am sticking with the previous version.

V11.3 does offer a little increase in performance but it come with a Energy demand increase that offset the increase in hashrate.
Overall, I was earning more/card/day a week ago

Ok great now try 11.4

[nirbo]

Hey everyone,

I am attempting to use the ETH miner v11.3 with NVIDIA V100 GPUs, however I keep getting the following error:

Code:

GPU 0, GpuMiner cu_kd failed 8 (0), invalid device function
GPU 0, Calc DAG failed!

I'm using Ubuntu 16.04 and I've tried with both CUDA 9.0 and 9.1

@Claymore, would you please advise?
How do I get this to work? to my understanding, v11 of the miner is supposed to support CUDA 9.1.


Thanks.

Hey @Claymore,

Any tips on what I may be doing wrong?
How can I get the miner to work with V100 GPUs?


Thank you!

[Omg9500]

You need to find the best settings for your own cards and memory type. My cards are a mix of Asus RX 580/480 with Samsung memory (first three with 2050 MHz memory overclock), Sapphire Nitro+ RX 580's with Samsung memory (2100 MHz memory overclock), Hynix memory (2200 MHz memory oveclock) and a MSI Gaming X RX 580 with Samsung memory (last card with a 2050 MHz memory overclock).

Code:

-cclock 1200,1150,1200,1200,1200,1200,1200,1200,1200,1200,1200,1200,1200
-mclock 2050,2050,2050,2200,2100,2100,2100,2220,2220,2220,2100,2220,2050
-powlim -10,-5,-10,-15,-10,-15,-10,-10,-5,-10,-10,-10,-10
-cvddc 950,950,950,850,900,850,850,900,900,900,850,900,900
-mvddc 950,950,950,850,900,850,850,900,900,900,850,900,900

Hello m8,
are these clocks and voltages for dual mining or solo ?
I have 6 of Sapphire rx 580 se with micron memory blue ones.
Nearly all of cards have up %75 asics quality. Whats the best values of clock mlock and dcc values for dual and solo eth mining ?
Im using -cvdcc 865,875,865,865,865,865 -cclock 1240,1240,1240,1240,1240,1240 -mclock 2250,2240,2245,2245,2250,2230 -mvdcc 875,875,880,875,880,880 and i don't use -powlim. Should i use powlim and these values is high or ok ? Hwinfo don't show errors. I also use -ethi 16. I want dual mining. Which values should i use both solo and dual mining ? I got 32.4 - 32.0 mh/s total is 193.4 mh/s. Hwinfo don't show errors but share is really so different. Sometimes calculated 350 mh/s sometimes 58 mh/s average on nanopool. Avarage values always under 190mh/s. I did 48 hours test ethermine and nano result same always 48 hours under 190 mh/s calculated.
Thanks for reply

I would say that largely depends on the BIOS... whether modified or not. I have a few of the same cards and got a modified BIOS from the guys over at anorak.tech and my settings in overdriventool are core 1175/850 and mem 2200/875 Hashrate per card 31.5 mh/s can't say powerdraw on the wall per cards but my rig with 13 cards incl. monitor draws 1750 watt from the wall.