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

[lecaillou]

Hi, I took Sia for second mining, who is the best? Decred/Siacoin/Lbry/Pascal???

[agis6]

The miner progressively shows/detects less and less memory for detected GPUs
When the miner starts it shows 8148 MB for each card, and it goes all the way down to 1000 MB per card. This can happen in an hour or in 12 hours. Stopping and restarting the miner fixes the problem.

OS: Ubuntu 16.04.1
AMD drivers: 16.30
All cards are ASUS ROG STRIX-RX480-8G-GAMING

Has anyone else experienced this?
@Claymore: Could it be that you are not releasing GPU memory between internal miner restarts?

[smaxz]

The miner progressively shows/detects less and less memory for detected GPUs
When the miner starts it shows 8148 MB for each card, and it goes all the way down to 1000 MB per card. This can happen in an hour or in 12 hours. Stopping and restarting the miner fixes the problem.

OS: Ubuntu 16.04.1
AMD drivers: 16.30
All cards are ASUS ROG STRIX-RX480-8G-GAMING

Has anyone else experienced this?
@Claymore: Could it be that you are not releasing GPU memory between internal miner restarts?

Memleak when it fucks up. Then it restarts itself to hog more memory with the second copy of the DAG.

yes, definitely sounds like a memory leak to me.

grab some paper towels and check your motherboard/ram slots particularly, hopefully its not so bad that you are leaking underneath the rig too.

once i had stacked a couple rigs, like bunk beds you know.. so i could utilize the same 24" box fan for 14 gpu's.. and the top rig memory leaked all over the rig underneath it.. huge mess.

who would have thought all that fluid fits in such small semiconductors?

man i wish ram just ran on smoke like cpu's do..

[deadsix]

The miner progressively shows/detects less and less memory for detected GPUs
When the miner starts it shows 8148 MB for each card, and it goes all the way down to 1000 MB per card. This can happen in an hour or in 12 hours. Stopping and restarting the miner fixes the problem.

OS: Ubuntu 16.04.1
AMD drivers: 16.30
All cards are ASUS ROG STRIX-RX480-8G-GAMING

Has anyone else experienced this?
@Claymore: Could it be that you are not releasing GPU memory between internal miner restarts?

Memleak when it fucks up. Then it restarts itself to hog more memory with the second copy of the DAG.

yes, definitely sounds like a memory leak to me.

grab some paper towels and check your motherboard/ram slots particularly, hopefully its not so bad that you are leaking underneath the rig too.

once i had stacked a couple rigs, like bunk beds you know.. so i could utilize the same 24" box fan for 14 gpu's.. and the top rig memory leaked all over the rig underneath it.. huge mess.

who would have thought all that fluid fits in such small semiconductors?

man i wish ram just ran on smoke like cpu's do..

Umm ... thats not what Wolf0 was referring to, memory leak means the program is unable to release memory it no longer is using causing memory usage to spike way above usual levels.
What you are talking about is thermal compound leaking oil or something.

[smaxz]

oh. you're probably right. my bad

[thaelin]

@pustul
    I currently am using GTX 1060's and I braved the over clock world. I can push mine to near 21 but just shy before I blue screen.  I have to just accept that 20.5 is solid and go with it. Yours will go a bit faster but don't get greedy as all you will get is an unstable machine. Been there. Flaky board is the reason rig01 keeps dumping. New one in the works.

@Claymore
   Many thanks for the S/W and your time. Single (eth) for now but going to go dual when rack goes live.

thay

[dragonmike]

Cards have bios Mod

And Settings from the OS are

Core 1050 / Mem frequency 1930 / Powerstage 2 --> with simplemining.net as OS i hate Windows

How much lower did you have to clock your mem compared to mining ETH only?

I'm asking because I need to apply my overclocks at bios level on my cards so I'm trying to gauge in advance to save time before I start flashing several cards over and over again...

[Adeel83]

Trying to get my rig stable. I'm using miningpoolhub, and the rig freezes when the connection is lost to the stratum server. It's happening from time to time, and the log always shows the same message right at the end:

socket was closed remotely
ETH: Connection lost, retry in 20 sec...
(*system freezes, hard reboot required*)

It mines for several hours (sometimes days) and crashes when the tcp connection is lost (the connection is bound to get dropped every now and then given the nature of tcp/ip).

Sometimes it just freezes right after the "socket was closed" message, and doesn't get to the connection lost message. Freezes the whole OS, which is super annoying. This is not an Internet or firewall issue, I mean, come on, it shouldn't freeze the machine when the connection is dropped.

Windows 10. 16.8.2. 6x red devil rx 480s, stock bios and clocks. Claymore 9.2, -gser 2. Temps running nice and cool, so it's not that.

What gives?!?

[AlainC]

I have heard of MESA drivers being the opensource alternative to AMD GPU PRO of ATI
http://www.phoronix.com/scan.php?page=article&item=amdgpu-pro-1710&num=1
http://www.omgubuntu.co.uk/2017/05/mesa-17-1-ubuntu-ppa

Is there any experience, and howto, with Claymore ? is it better ?
Does ethdcrminer64 works better with MESA ? (I think of ADL non support by amdgpu pro v17)

Nb: I can make a comparison on my 2xRX480 is someone point me to a howto... but It seems hard to rollback (seen some old howto that imply changing kernel).

[agis6]

The miner progressively shows/detects less and less memory for detected GPUs
When the miner starts it shows 8148 MB for each card, and it goes all the way down to 1000 MB per card. This can happen in an hour or in 12 hours. Stopping and restarting the miner fixes the problem.

OS: Ubuntu 16.04.1
AMD drivers: 16.30
All cards are ASUS ROG STRIX-RX480-8G-GAMING

Has anyone else experienced this?
@Claymore: Could it be that you are not releasing GPU memory between internal miner restarts?

Memleak when it fucks up. Then it restarts itself to hog more memory with the second copy of the DAG.

yes, definitely sounds like a memory leak to me.

grab some paper towels and check your motherboard/ram slots particularly, hopefully its not so bad that you are leaking underneath the rig too.

once i had stacked a couple rigs, like bunk beds you know.. so i could utilize the same 24" box fan for 14 gpu's.. and the top rig memory leaked all over the rig underneath it.. huge mess.

who would have thought all that fluid fits in such small semiconductors?

man i wish ram just ran on smoke like cpu's do..

Nice. You definitely have a lot of time to spare... Wish I did, too

The question still stands: Has anyone else experienced this?
@Claymore: Do you need any more info to check this?

[usuksobad]

Hey guys,

Wich -mport combo should i use to be able to control gpus afterburner through awesome miner, kinda dont get it. Tried 3333, -3333, 0. nothing helps. I hope i wont need to overclock every rig manually.

Thanks

[newmz]

@pustul
    I currently am using GTX 1060's and I braved the over clock world. I can push mine to near 21 but just shy before I blue screen.  I have to just accept that 20.5 is solid and go with it. Yours will go a bit faster but don't get greedy as all you will get is an unstable machine. Been there. Flaky board is the reason rig01 keeps dumping. New one in the works.

@Claymore
   Many thanks for the S/W and your time. Single (eth) for now but going to go dual when rack goes live.

thay

I have 2 1060's and 2 1070's which are just mining ETH.

The 1060's (6GB) are set to core -200 mem +750 power 94% and they hash at 22.6 consistently.

The 1070's are similar settings but power at 84% and they do 30-31mh/s.

whole rig using around 600W.

[newmz]

Hi guys,

I know NVIDIA cards are not ideal for eth mining but what should I expect from a gtx 1080? I have latest drivers (382.05) on win10 anniversary and I only get about 25 Mh/s even with oc.

1070s and 1060s are great for eth mining. The 1080 not so much.

See my post above, but my 1060's are doing 22.6mh/s and 1070's 30-31mh/s.

2 of each card on asrock h81 pro btc - 600w 108mh/s.

Using Windows 8.1 - I believe some of the recent updates with Windows 10 have messed with ETH mining. With Win8.1 you can find your best driver and KEEP it instead of having Windows 10 ramming the latest driver into your machine without your consent.... like driver rape or something.

[f00ch0w]

I've set Dual-Mining mode, edited the claymore.stub config file and it keeps rejecting shares from Decred... What did I do wrong? Did I maybe put wrong values or edited something I shouldn't in the claymore.stub file? Can someone give me a guide step by step what to do?

[smaxz]

Nice. You definitely have a lot of time to spare... Wish I did, too

The question still stands: Has anyone else experienced this?
@Claymore: Do you need any more info to check this?

only while running windows with sub par memory specs (2gig ram)

you may possibly have run into a system compatibility issue? i would take your dmesg and any logged system errors to the amdgpu-pro mailing list.

possibly also consider using the pure open source driver stack individually instead of the entire amdgpu-pro package if you are up for that kind of thing.

good luck!

[chohav]

Hi guys,

I know NVIDIA cards are not ideal for eth mining but what should I expect from a gtx 1080? I have latest drivers (382.05) on win10 anniversary and I only get about 25 Mh/s even with oc.

1070s and 1060s are great for eth mining. The 1080 not so much.

See my post above, but my 1060's are doing 22.6mh/s and 1070's 30-31mh/s.

2 of each card on asrock h81 pro btc - 600w 108mh/s.

Using Windows 8.1 - I believe some of the recent updates with Windows 10 have messed with ETH mining. With Win8.1 you can find your best driver and KEEP it instead of having Windows 10 ramming the latest driver into your machine without your consent.... like driver rape or something.

Good job newmz, those are nice numbers. I am still AMD all the way, but thinking of building one Nvidia rig, and the numbers from the 1070s you reached are really good. What is the consumption from the wall on the 1070, which model do you use and have you tried dual with? Thanks

[qqqq]

Can anyone confirm that 9.3 finds a blocks in SOLO ? 650 MH\s and 11 days still no block. Bad luck ?

Claymore miner is designed to mine on pools only. "Claymore' Solo" means ETH only, "Claymore' Duo" means ETH + another coin.
You have to check your earnings on the pool you are mining to. If you can share your config and bat files we might be able to point you to the right direction.

Are you newbie or what ? I'm in solo direct to the wallet since ver 5.0 or so (maybe wrong, forgot when exactly he implemented solo for ETH).

Sometimes Claymore did the mistakes in his new versions and it do not finds any blocks. He confirmed that bugs, it was about half or year ago. So maybe it happens again. It maybe a bad luck but anyway.

So if anyone in solo mode (direct to the geth) with 9.3 could you please confirm that it finds any blocks. Thanks

[Adeel83]

Trying to get my rig stable. I'm using miningpoolhub, and the rig freezes when the connection is lost to the stratum server. It's happening from time to time, and the log always shows the same message right at the end:

socket was closed remotely
ETH: Connection lost, retry in 20 sec...
(*system freezes, hard reboot required*)

It mines for several hours (sometimes days) and crashes when the tcp connection is lost (the connection is bound to get dropped every now and then given the nature of tcp/ip).

Sometimes it just freezes right after the "socket was closed" message, and doesn't get to the connection lost message. Freezes the whole OS, which is super annoying. This is not an Internet or firewall issue, I mean, come on, it shouldn't freeze the machine when the connection is dropped.

Windows 10. 16.8.2. 6x red devil rx 480s, stock bios and clocks. Claymore 9.2, -gser 2. Temps running nice and cool, so it's not that.

What gives?!?

I upgraded to Claymore 9.3. Seems to be holding stable on socket closed's so far. We'll see.

BTW since we're playing Claymore recurring fees. Does s/he provide technical support?

[h!ghC]

Hello,

i have a problem: when I start mining (only ETH: -mode 1) at the suprnova.cc pool everything is fine for the first 2-3 minutes. But then I get no more hashes (no "green lines". But it says: Total speed: 25Mh/s, but total shares: 0. Any idea why?

my config:

-epool stratum+tcp://eth.suprnova.cc:5005
-ewal username.eth1
-epsw password
-dpool stratum+tcp://dcr.suprnova.cc:3252
-ewal username.dcr1
-epsw password
-mode 1
-allpools 1

[Pustul]

@pustul
    I currently am using GTX 1060's and I braved the over clock world. I can push mine to near 21 but just shy before I blue screen.  I have to just accept that 20.5 is solid and go with it. Yours will go a bit faster but don't get greedy as all you will get is an unstable machine. Been there. Flaky board is the reason rig01 keeps dumping. New one in the works.

@Claymore
   Many thanks for the S/W and your time. Single (eth) for now but going to go dual when rack goes live.

thay

I have 2 1060's and 2 1070's which are just mining ETH.

The 1060's (6GB) are set to core -200 mem +750 power 94% and they hash at 22.6 consistently.

The 1070's are similar settings but power at 84% and they do 30-31mh/s.

whole rig using around 600W.

Yeah that's what I'm talking about, how come my 1080 is doing only 25mh/s when your 1070s are doing 30-31mh/s? Are you using windows or linux? Which drivers? Thanks