Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1328. (Read 6590565 times)

Thanks!

How to make this16gb virtual memory?   And where?
Could it be reason for low  280/370/380 series  amd, hashrate?


Ok I found but what  consequences could  have to small size?

awesome--will do.  thanks!

Was fixed a few days ago, re-download v4.1 or use 4.2.

completely useless bug report follows, but I feel obligated to offer it.

4.1, dual mining in linux.  after several days,

read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable
read(): Resource temporarily unavailable

attempts to restart mining just hang.
soft (AHCI) reboot fixes it (implies not a hard hang at GPU/driver).

I think you would see an accepted share in Claymore's miner for each found block. That's been my experience so far with Expanse/gexp.

EDIT: I can also confirm that v4.2 works for Expanse solo mining.

Thanks for the quick reply Claymore.

I think it may be something different.

Your readme file advises that for multi-gpu support 16gb is needed as virtual memory.
I think I saw this we the program ran also. I went to change the virtual memory in advanced settings.
However, the virtual memory never saved properly. I set the memory through command prompt also.
The memory allocation either never saved properly or windows came back with an error after rebooting.
It seems like Virtual memory cannot be set onto USB drives, which is how I have been running Windows.

I hope that this was the problem all along, which is easily fixed by going to the shop and picking up a bit of hardware.
Assuming this was the problem, then you still have awesome software, though 'probably' cannot be run on USB (power saving?).
I like this software because it just hashes better. Perhaps in future an option could be coded to create DAG files again.

I have to use the MSI afterburner to control the voltage and frequency. So I have to use the 4.2 beta 2.

By default (no "-tt" option or "-tt 1") v4.2 Beta resets fans management to default, i.e. AMD drivers will manage fans. So it's ok if you don't use Afterburner to manage fans, but will cause a problem if Afterburner is used.
v4.2 Beta2 does not reset fans management so Afterburner can manage fans.

If I do not use the fan speed options, do I need to upgrade to the V4.2 Beta 2? Or what is the issue with fans?

v4.2 Beta2: Fixed issue with fans.

About wrong GPU indexes: I explained it a few times, now added to "KNOWN ISSUES" in OP.

Yes. I can also confirm this.

Try the new "serialization" feature in 4.2
Check risers and PSU... PSUs are supposed to "age" 5% per year...

"clEnqueueNDRangeKernel" function returns "CL_MEM_OBJECT_ALLOCATION_FAILURE" error from your log. It is something system-related. Actually this question is to AMD drivers developers team, not to me, I have no ideas why OpenCL cannot reserve GPU memory for DAG on your system, sorry.

Hi Claymore,

Testing v4.2 now with solo mode

1. 5 x 390 @ 148MHs x 4 rigs = 592MHs total hash pointing to standalone pc in same LAN running Geth (AMD 8 core, 16GB RAM, 120GB SSD)
2. Geth is fully synchronised and showing current block #
3. AMD PC is also running CDMv42 on 1 single 390.
4. All 4 rigs are hashing and communicating with Geth at 29.7MHs per card or 148.5MHs per rig.
5. All 4 rigs are showing GPU temps.
6. CDMv42 on a single GPU is NOT showing GPU temps.
7. All rigs are on Catalyst 16.3.2, and 16GB-20GB virtual memory, W8.1, wired LAN connection.
8. DevFee mining is active (showing you are mining at dwarfpool EU)

Questions:
1. I have used other proxy solution before this, and usually ethminer sends shares to Geth via a proxy server, and when block is solved it is accepted by Geth... since this version v4.2 is directly hashing into geth, the only indication that I successfully mined a solo block is on the CDM console or local MIST wallet, correct?
2. When you start Geth, what is your syntax? Do you use any switches? I started Geth with console mode.

Let me know if I have missed any steps. I have a feeling my Geth is not working correctly (because there is no shares accepted after running for an hour) or maybe I am just impatient because this solo mining after all.

Thanks

I will be  very  apreciate for  28x and 3xx series optymalization too

Claymore, do you have any advice on this please? I have 20 R9 280xs that would love to start mining with your program

In CDM FAQ:

KNOWN ISSUES

- Weak/old cards like 7xxx/270/270X cannot handle dual mining properly, Ethereum mining is slower by about 5%.

i'm having trouble having peers in my ETH wallet...any add nodes or something like that?

anyone using 4.1 having this issue?

08:42:40:295   a50   Cannot resolve 'stratum-dcr-sg.coinmine.pl'
08:42:40:295   a50    DCR: Stratum - Cannot connect to stratum-dcr-sg.coinmine.pl:2222
08:42:40:295   a50    DCR: Stratum - Failed to connect, retry in 20 sec...
08:43:00:298   a50   Cannot resolve 'stratum-dcr-sg.coinmine.pl'
08:43:00:299   a50    DCR: Stratum - Cannot connect to stratum-dcr-sg.coinmine.pl:2222
08:43:00:300   a50    DCR: Stratum - Failed to connect, retry in 20 sec...
08:43:01:218   bf0   ETH: Stratum - Cannot connect to eth-hk.dwarfpool.com:8008
08:43:01:220   bf0   ETH: Stratum - Failed to connect, retry in 20 sec...

Btw @Claymore, I have realized that if you have issues with the Eth connection, even if the Dcr is connected, it will not do any work at all until your Eth connection is connected. Is this suppose to be worked out like that or was it overlooked? hmm