CUDA error 11-cannot write buffer for DAG

[eniacs]

Had this same problem. Started to think like the others above, "Oh no i cant mine anymore as dag files too big for my GPU" etc.

Take DAG file size, multiply by number of GPUS and set virtual memory to this size.

For me;
12x3GB GPU's
12x2.3GB = 27.6 GB so set virtual memory to 32GB (bit of overhead for DAG file size increases).

Done.

[1715003661]

1715003661

[1715003661]

1715003661

[1715003661]

1715003661

[xtom1050]

Had this same problem. Started to think like the others above, "Oh no i cant mine anymore as dag files too big for my GPU" etc.


 
Take DAG file size, multiply by number of GPUS and set virtual memory to this size.

For me;
12x3GB GPU's
12x2.3GB = 27.6 GB so set virtual memory to 32GB (bit of overhead for DAG file size increases).

Done.

Unfortunately it does not help, 6 cards with 30Gb virtual memory, still the same error.

[eniacs]

Something else must be wrong. Im still mining on my 3GB's which are a mix of different brands, they are nothing special.

Try reinstalling drivers? Got another pc you could try one GPU in to confirm the problem is with that rig?

[dzeb]

Same problem, mining with claymore and 3gb 1060, no problem with ETH, only with ETC.

[xtom1050]

Something else must be wrong. Im still mining on my 3GB's which are a mix of different brands, they are nothing special.

Try reinstalling drivers? Got another pc you could try one GPU in to confirm the problem is with that rig?

Got another rig mixed 1050s+1060s, same problem. Friend has 16 1060s in 2 rigs, same problem
All 3GB, different brands, different driver versions.

[bitcoin.dbi]

Got the same problem all day long today on 2 rigs... FIxed it with the help of rasers mixing. No idea why but it helped....
1063. 1066. 1050 ti.
Doesnt metter.... All got the same problem. AMD worked fine..

[Ahorradorr]

good morning

friends tell them what I use
I have working 8 GPU 1060 3gb and today I present the problem cuda 11. Thanks to the -eres command 0 solve it.

For reference I have working with the 382.33 msi driver on the msi z170a board

Happy day

thanks for the solution

[xtom1050]

good morning

friends tell them what I use
I have working 8 GPU 1060 3gb and today I present the problem cuda 11. Thanks to the -eres command 0 solve it.

For reference I have working with the 382.33 msi driver on the msi z170a board

Happy day

thanks for the solution

Happy you Good to hear it is solved for someone, this command solved me once, the stopped miner and since restarting it has been doing nothing, getting error. Same applies for lidag or allcoins etc commands, they do not help

[WolfH]

I have 2x 1060 3GB

ETC dont work on WIN 10
ETH work on WIN 10
but...

ETC and ETH work on LINUX

i think is WIN 10 problem. He take 200mb or more from GPU. Meaby install WIN 7 help.

good luck comrades

[cuchuoivts]

take the command -eres 0 in the claymores bat file and you will be happy for sometimes more

Thank for your information,I have added -eres 0 into end of my config file --> 3PC are working normal now (My card is MSI GTX 1060 3GB)
-epool etc-asia1.nanopool.org:19999
-ewal 0xxxxxxxxxxxx
-eworker testing
-epsw x
-eres 0

That great!

[WolfH]

take the command -eres 0 in the claymores bat file and you will be happy for sometimes more

Thank for your information,I have added -eres 0 into end of my config file --> 3PC are working normal now (My card is MSI GTX 1060 3GB)
-epool etc-asia1.nanopool.org:19999
-ewal 0xxxxxxxxxxxx
-eworker testing
-epsw x
-eres 0

That great!

What system you use?

[pigseye]

Trying to mine ETC in Windows 10 (version 1709) with Gigabyte 1060 3gb and no luck with Claymore or Phoenix miners with -eres 0 setting and Nvidia drivers 390.65.  This is on the ethteam pool.

Going to roll back drivers to 382.33 to see what happens.


I can mine Ethereum in Claymore with the -eres 0 setting in the suprnova pool so having an issue with Etc is surprising.

Does the pool make a difference?  I don't see how it could.

Any other suggestions?

Edit - RESOLVED!

Rolling back to 382.33 drivers fixed the issue!  So at least for a little while longer I can mine ETC on 3GB 1060s

[ansuz]

I have same problem.
I had to restarted miners (2 miner) and after that i can't mining ETC.

Specification:
Win10 Anniversary
Virtual cash 20000 to 100000 (I tryed fix on 50000).
MSI gtx 1060 3GB v. driver 388.43.
I tried Claymore (-eres 0 doesn't work for me), Ethminer 13 and Phoenix 2.6 and still writen same Out of memory.

Interesting is that phoenix alcating DAG 2,41G and good for epocho up to 181.
Here says to is good: https://investoon.com/tools/dag_size


17574:16:33:17.481: eths Eth: New job #4276d251 from eu1.nanopool.org:19999; diff: 10000MH
17574:16:33:17.483: GPU1 GPU1: Starting up...
17574:16:33:17.483: GPU1 Eth: Generating light cache for epoch #179
17574:16:33:17.494: GPU2 GPU2: Starting up...
17574:16:33:17.499: GPU3 GPU3: Starting up...
17574:16:33:17.500: GPU4 GPU4: Starting up...
17574:16:33:17.503: GPU5 GPU5: Starting up...
17574:16:33:17.537: eths Eth: Received: {"jsonrpc":"2.0","id":0,"result":["0x4276d25191bf88d687eee43f2ba8b57a8290df5d578e4c063222181fb07fb2a9","0xa74dd84414948aba4bfd6580908dbf9c32a494aa336609746d19a35bb4cccb24","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}
17574:16:33:20.693: GPU2 GPU2: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.718: GPU4 GPU4: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.750: GPU3 GPU3: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.830: GPU4 GPU4: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU4 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU4 GPU4 initMiner error: out of memory
17574:16:33:20.831: GPU2 GPU2: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU2 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU2 GPU2 initMiner error: out of memory
17574:16:33:20.831: GPU3 GPU3: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.832: GPU3 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.832: GPU3 GPU3 initMiner error: out of memory
17574:16:33:20.894: GPU5 GPU5: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.986: GPU5 GPU5: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.986: GPU5 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.986: GPU5 GPU5 initMiner error: out of memory
17574:16:33:21.122: GPU1 GPU1: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:21.209: GPU1 GPU1: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:21.210: GPU1 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:21.210: GPU1 GPU1 initMiner error: out of memory
17574:16:33:21.718: wdog Thread(s) not responding. Restartin


WTF? Any idea? What hapen? How it fix it?

[pigseye]

Try 382.33.  it worked for me.

I have same problem.
I had to restarted miners (2 miner) and after that i can't mining ETC.

Specification:
Win10 Anniversary
Virtual cash 20000 to 100000 (I tryed fix on 50000).
MSI gtx 1060 3GB v. driver 388.43.
I tried Claymore (-eres 0 doesn't work for me), Ethminer 13 and Phoenix 2.6 and still writen same Out of memory.

Interesting is that phoenix alcating DAG 2,41G and good for epocho up to 181.
Here says to is good: https://investoon.com/tools/dag_size


17574:16:33:17.481: eths Eth: New job #4276d251 from eu1.nanopool.org:19999; diff: 10000MH
17574:16:33:17.483: GPU1 GPU1: Starting up...
17574:16:33:17.483: GPU1 Eth: Generating light cache for epoch #179
17574:16:33:17.494: GPU2 GPU2: Starting up...
17574:16:33:17.499: GPU3 GPU3: Starting up...
17574:16:33:17.500: GPU4 GPU4: Starting up...
17574:16:33:17.503: GPU5 GPU5: Starting up...
17574:16:33:17.537: eths Eth: Received: {"jsonrpc":"2.0","id":0,"result":["0x4276d25191bf88d687eee43f2ba8b57a8290df5d578e4c063222181fb07fb2a9","0xa74dd84414948aba4bfd6580908dbf9c32a494aa336609746d19a35bb4cccb24","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}
17574:16:33:20.693: GPU2 GPU2: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.718: GPU4 GPU4: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.750: GPU3 GPU3: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.830: GPU4 GPU4: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU4 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU4 GPU4 initMiner error: out of memory
17574:16:33:20.831: GPU2 GPU2: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU2 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU2 GPU2 initMiner error: out of memory
17574:16:33:20.831: GPU3 GPU3: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.832: GPU3 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.832: GPU3 GPU3 initMiner error: out of memory
17574:16:33:20.894: GPU5 GPU5: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.986: GPU5 GPU5: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.986: GPU5 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.986: GPU5 GPU5 initMiner error: out of memory
17574:16:33:21.122: GPU1 GPU1: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:21.209: GPU1 GPU1: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:21.210: GPU1 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:21.210: GPU1 GPU1 initMiner error: out of memory
17574:16:33:21.718: wdog Thread(s) not responding. Restartin


WTF? Any idea? What hapen? How it fix it?

[nitrobg]

Can anyone try mining on Linux, it might solve the issue?

[ansuz]

Try 382.33.  it worked for me.

I have same problem.
I had to restarted miners (2 miner) and after that i can't mining ETC.

Specification:
Win10 Anniversary
Virtual cash 20000 to 100000 (I tryed fix on 50000).
MSI gtx 1060 3GB v. driver 388.43.
I tried Claymore (-eres 0 doesn't work for me), Ethminer 13 and Phoenix 2.6 and still writen same Out of memory.

Interesting is that phoenix alcating DAG 2,41G and good for epocho up to 181.
Here says to is good: https://investoon.com/tools/dag_size


17574:16:33:17.481: eths Eth: New job #4276d251 from eu1.nanopool.org:19999; diff: 10000MH
17574:16:33:17.483: GPU1 GPU1: Starting up...
17574:16:33:17.483: GPU1 Eth: Generating light cache for epoch #179
17574:16:33:17.494: GPU2 GPU2: Starting up...
17574:16:33:17.499: GPU3 GPU3: Starting up...
17574:16:33:17.500: GPU4 GPU4: Starting up...
17574:16:33:17.503: GPU5 GPU5: Starting up...
17574:16:33:17.537: eths Eth: Received: {"jsonrpc":"2.0","id":0,"result":["0x4276d25191bf88d687eee43f2ba8b57a8290df5d578e4c063222181fb07fb2a9","0xa74dd84414948aba4bfd6580908dbf9c32a494aa336609746d19a35bb4cccb24","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}
17574:16:33:20.693: GPU2 GPU2: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.718: GPU4 GPU4: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.750: GPU3 GPU3: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.830: GPU4 GPU4: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU4 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU4 GPU4 initMiner error: out of memory
17574:16:33:20.831: GPU2 GPU2: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU2 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU2 GPU2 initMiner error: out of memory
17574:16:33:20.831: GPU3 GPU3: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.832: GPU3 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.832: GPU3 GPU3 initMiner error: out of memory
17574:16:33:20.894: GPU5 GPU5: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.986: GPU5 GPU5: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.986: GPU5 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.986: GPU5 GPU5 initMiner error: out of memory
17574:16:33:21.122: GPU1 GPU1: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:21.209: GPU1 GPU1: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:21.210: GPU1 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:21.210: GPU1 GPU1 initMiner error: out of memory
17574:16:33:21.718: wdog Thread(s) not responding. Restartin


WTF? Any idea? What hapen? How it fix it?

SOLVED
Yes, great man, thank you so much.

Works with version 382.33 and mining program Phoenix 2.6, Ethminer 12 (13 does not supported old driver), I did not try Claymore.

Do you think it's the intention of Nvidia?

[marvel213]

Try 382.33.  it worked for me.

I have same problem.
I had to restarted miners (2 miner) and after that i can't mining ETC.

Specification:
Win10 Anniversary
Virtual cash 20000 to 100000 (I tryed fix on 50000).
MSI gtx 1060 3GB v. driver 388.43.
I tried Claymore (-eres 0 doesn't work for me), Ethminer 13 and Phoenix 2.6 and still writen same Out of memory.

Interesting is that phoenix alcating DAG 2,41G and good for epocho up to 181.
Here says to is good: https://investoon.com/tools/dag_size


17574:16:33:17.481: eths Eth: New job #4276d251 from eu1.nanopool.org:19999; diff: 10000MH
17574:16:33:17.483: GPU1 GPU1: Starting up...
17574:16:33:17.483: GPU1 Eth: Generating light cache for epoch #179
17574:16:33:17.494: GPU2 GPU2: Starting up...
17574:16:33:17.499: GPU3 GPU3: Starting up...
17574:16:33:17.500: GPU4 GPU4: Starting up...
17574:16:33:17.503: GPU5 GPU5: Starting up...
17574:16:33:17.537: eths Eth: Received: {"jsonrpc":"2.0","id":0,"result":["0x4276d25191bf88d687eee43f2ba8b57a8290df5d578e4c063222181fb07fb2a9","0xa74dd84414948aba4bfd6580908dbf9c32a494aa336609746d19a35bb4cccb24","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}
17574:16:33:20.693: GPU2 GPU2: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.718: GPU4 GPU4: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.750: GPU3 GPU3: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.830: GPU4 GPU4: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU4 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU4 GPU4 initMiner error: out of memory
17574:16:33:20.831: GPU2 GPU2: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU2 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU2 GPU2 initMiner error: out of memory
17574:16:33:20.831: GPU3 GPU3: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.832: GPU3 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.832: GPU3 GPU3 initMiner error: out of memory
17574:16:33:20.894: GPU5 GPU5: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.986: GPU5 GPU5: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.986: GPU5 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.986: GPU5 GPU5 initMiner error: out of memory
17574:16:33:21.122: GPU1 GPU1: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:21.209: GPU1 GPU1: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:21.210: GPU1 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:21.210: GPU1 GPU1 initMiner error: out of memory
17574:16:33:21.718: wdog Thread(s) not responding. Restartin


WTF? Any idea? What hapen? How it fix it?

SOLVED
Yes, great man, thank you so much.

Works with version 382.33 and mining program Phoenix 2.6, Ethminer 12 (13 does not supported old driver), I did not try Claymore.

Do you think it's the intention of Nvidia?

Working for me as well.  Using claymore dual mine.  Installed drivers and you have to add -eres 0 in the bat file.

It won't be long before the 3gb card won't be able to mine ETH and ETC however. 

[xtom1050]

Try 382.33.  it worked for me.

I have same problem.
I had to restarted miners (2 miner) and after that i can't mining ETC.

Specification:
Win10 Anniversary
Virtual cash 20000 to 100000 (I tryed fix on 50000).
MSI gtx 1060 3GB v. driver 388.43.
I tried Claymore (-eres 0 doesn't work for me), Ethminer 13 and Phoenix 2.6 and still writen same Out of memory.

Interesting is that phoenix alcating DAG 2,41G and good for epocho up to 181.
Here says to is good: https://investoon.com/tools/dag_size


17574:16:33:17.481: eths Eth: New job #4276d251 from eu1.nanopool.org:19999; diff: 10000MH
17574:16:33:17.483: GPU1 GPU1: Starting up...
17574:16:33:17.483: GPU1 Eth: Generating light cache for epoch #179
17574:16:33:17.494: GPU2 GPU2: Starting up...
17574:16:33:17.499: GPU3 GPU3: Starting up...
17574:16:33:17.500: GPU4 GPU4: Starting up...
17574:16:33:17.503: GPU5 GPU5: Starting up...
17574:16:33:17.537: eths Eth: Received: {"jsonrpc":"2.0","id":0,"result":["0x4276d25191bf88d687eee43f2ba8b57a8290df5d578e4c063222181fb07fb2a9","0xa74dd84414948aba4bfd6580908dbf9c32a494aa336609746d19a35bb4cccb24","0x000000006df37f675ef6eadf5ab9a2072d44268d97df837e6748956e5c6c2116"]}
17574:16:33:20.693: GPU2 GPU2: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.718: GPU4 GPU4: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.750: GPU3 GPU3: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.830: GPU4 GPU4: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU4 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU4 GPU4 initMiner error: out of memory
17574:16:33:20.831: GPU2 GPU2: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.831: GPU2 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.831: GPU2 GPU2 initMiner error: out of memory
17574:16:33:20.831: GPU3 GPU3: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.832: GPU3 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.832: GPU3 GPU3 initMiner error: out of memory
17574:16:33:20.894: GPU5 GPU5: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:20.986: GPU5 GPU5: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:20.986: GPU5 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:20.986: GPU5 GPU5 initMiner error: out of memory
17574:16:33:21.122: GPU1 GPU1: Allocating DAG (2.41) GB; good for epoch up to #181
17574:16:33:21.209: GPU1 GPU1: Allocating light cache buffer (38.6) MB; good for epoch up to #181
17574:16:33:21.210: GPU1 CUDA error in CudaProgram.cu:301 : out of memory (2)
17574:16:33:21.210: GPU1 GPU1 initMiner error: out of memory
17574:16:33:21.718: wdog Thread(s) not responding. Restartin


WTF? Any idea? What hapen? How it fix it?

Thank you very very much mate, this driver works with Claymore 9.7

[Ahorradorr]

FRIENDS HELP

I have a problem with my rig, start the claymore and just 1 hour after the uninhibited machine. I have to restart it abruptly.

They will have some solution.

a few days ago present problems with the error cuda 11

the solvent with the -eres 0 but after an unexpected shutdown in the machine now I present this fault.

I have claymore 9.8
msi 1060 3gb
windows 10
 
I'm mining ETC

[ansuz]

FRIENDS HELP

I have a problem with my rig, start the claymore and just 1 hour after the uninhibited machine. I have to restart it abruptly.

They will have some solution.

a few days ago present problems with the error cuda 11

the solvent with the -eres 0 but after an unexpected shutdown in the machine now I present this fault.

I have claymore 9.8
msi 1060 3gb
windows 10
 
I'm mining ETC

Yes problem is back, today one rig shut down and not possible start rig. Same problem. Have you any solution? 382.33 doesn´t work now. Phenix, Ethminer nothing. WTF?