Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1087. (Read 6590565 times)

I sometimes get a hang if the driver crash on one of the cards in the rig. How can I restart the miner f.ex every 3 hours? or more? Did you make a restart parameter?

I got this precise issue when my swap file was too small. Random gpu wasn't compiling dag while others where okay (up to 2 gpu failed depending of swap size)
Looking @ your rig's specs, I misread you had 4 Gb :/

AFAIR, I tried -gser with no success. I think this option serializes OpenCL code compilation (miner writes "POOL/SOLO version" and builds the code, thus delay). Then it runs the same DAG build process on all GPUs at the same time, and then a memory size should not make any sense.

I forgot to manually increase virtual RAM size, but I have 8G RAM installed and think that at least the same amount was added by paging, so it should not be a problem.

Ram / swap file size issues on w10
Either serialize cards init (-gser option) or add some ram or check real swap file size

Anybody runs CDM with 6 RX470 cards? I've been trying, still no success.

On Linux single card works fine, 2+ are visible but give weird hash rates (too low and unstable).
On W7 only 4 cards work, with 6-card mod ANY 5 of 6 cards work, one gives error code 43. I can change it to other, but up to 5 working.
On W8 5 of 6 cards work without any mods, one gives error. Seems the same issue as on W7 with 6-card mod.
On W10 all 6 cards are visible, but CDM shows errors creating DAG on 2-4 cards (different every time), -lidag does not help.

CDM 7.2, amdgpu pro on Linux, Win drivers 16.9.2, ASRock H81 Pro BTC, Sapphire RX470 Nitro+ 4GB, PSU 1300W, different risers (some strips and some USB ones, while I am waiting for a pack of 6 USB risers).

W7 and W8 work but seems like insufficient PCI-E resources to run 6 cards. Starting the 6th card initializes it and immediately gives code 43. ANY of 6.... So no hardware issues, but something else. Mining on any 5 cards is stable.

W10 and Linux - I have no idea what happens at all.

I had the same thing if i use remote connect and then starting Claymore. But when i connect using TeamViewer and restarts te miner it shows all the cards temp and fans speeds

Here's one of the original reviews for the 6990 - I like this line: "the total number of stream processors inside this 12in-long card is a staggering 3,072"  

http://www.bit-tech.net/hardware/2011/03/08/amd-radeon-hd-6990-review/3

And, I see some of them shipped with double lifetime warranties - I guess, someone's grandchildren are in for a real GPU treat

Anyone have the same issue ? .

I must give props to Claymore for his work on the older GPUs.

Radeon 6990
PreGCN. Since its not GCN 1.0 like Tahiti/Pitcairn it doesn't slow down due to DAG file increase.


Almost 6 years old. Still murdering it out at 26MH/s. 40nm technology. Loves running at 90C. And After 5 years the fan still is original.

This GPU probably made people the most money. Was the fastest GPU at the time for mining Bitcoins. Would of made millionaires during the BTC mining days. Made great money during LTC rush. Sucked at X11 unfortunately due to bad coding of the bins but picked it up with ETHEREUM mining again.

It's probably swapping ... try -r 720 to restart every 12 hours automatically

-r   Restart miner mode. "-r 0" (default) - restart miner if something wrong with GPU. "-r -1" - disable automatic restarting. -r >20 - restart miner if something
   wrong with GPU or by timer. For example, "-r 60" - restart miner every hour or when some GPU failed.

We're getting an issue on our linux-based NFS netbooted cluster with v7.2 in which after around 36 hours of operation, the miner starts to do a ridiculous amount of "disk" accesses. This saturates our gigabit ethernet (112 MiB/s) for about 5-10 minutes per rig.

As a workaround, we can restart the machines every 24 hours. We'd like to be able to do this with a cron job. What data should we send to the miner, and at which port, in order to make it restart?

Claymore ar you will make a miner for zcash?

Have a look at bios mod threads, if gpu vcore is too high, then memory voltage stays at 1 V whatever value you set. I don't remember the exact rule

The MSIs have slower memory, I think 6.7GHz. 7GHz 4GB cards do 22MH, 8GB 8GHz cards do ~24.3MH.  The MSI 470 8GB is a terrible card for mining  Has 6.6-6.7GHz memory, yet costs the same as a Sapphire 8GB with 8GHz memory

This is an option but its risky and I also think that it can void your warranty, I would not recommend you to do this honestly.

When you don't see change restart miner or reboot computer.
But if you want to see high hash rates (27-28mh) you MUST mod the bios.. there's a ton of info on it:
http://forum.ethereum.org/discussion/8845/rx-470-mining/p1

So, I have been using CDM 7.2 for about 5 days now on my W7 system. I was finally able to reduce my wattage pull through Wattman setting (and disabling dual mining), but am still stuck on a paltry 20.75 Mh/s on ETH per each of my MSI RX 470's.  I have tried everything on the CDM side, such as setting ETH intensity to the max of 16 in my command line, adding in all 5 of the recommended setx lines, set my virtual memory to 24GB, etc. None of this moved the needle at all.

So, I am now trying the overclocking route before getting into modding the bios on these.  A few things are perplexing though. In my Wattman, each of the 4 GPU's can only be increased from 1650 default to 1700.  These seems like a very low limit to me and setting all 4 to 1700 had zero impact.

So, using Afterburner, it actually gave me the option to raise Memory Clock all the way to 2100.  I only tried 1850 and applied to all 4 GPUs but yet again, absolutely NO impact... all 4 still mining at exactly 20.75. So, my question is, am I doing something wrong?  Shouldn't raising the the clock speed on these yield at least a 5-10% bump in mining hash-rate?  Or am I completely off base here?

I know of many people using these cards that have achieved 23+ right out of the box, so still trying to figure out why I cannot get close to that mark.

SIA is a slow coin. It takes a long time for each block to be confirmed. Your SIA coin will take awhile to show up. Confirmations can take 18-24 hours. I don't remember exactly but I know that my coin I mined today wouldn't show until tomorrow in most cases.

According to Suprnova:
Block time is 10 minutes for sia and difficulty is very high, it takes some time before you see some funds in your account.

Someone that actually mines SIA currently can probably give you more info. I haven't mined it for awhile.

When switching pools, is it possible to stay on alternate pool until main pool is actually up?  My main sia pool is down, but it keeps stopping mining on alternate pool to check, so there is repeated downtime.  Can this be done in a future version?

Because as I wrote above, GPU frequency/voltage for P1-P7 is read from modded BIOS and used. But for memory only P1 frequency was set, but voltage still is default 1000mV instead of flashed value. I don't know why.