Author

Topic: [Cast XMR] high speed XMR/CryptoNight miner for RX Vega GPUs (2 KHash/s) - page 105. (Read 206404 times)

newbie
Activity: 7
Merit: 0
So I have been able to get this running on my 11 4GB XFX RX 470s, I have one 6 card system and one 5 card system. On both I seem to get about 9% of my shares rejected because of "out of date job change". Is there no resolution to this? 9% is not a small amount either. This is well after 24 hours of straight mining. I get an 865 hash rate from each card 9% of that times 11 is over 870 H\s. In a sense I would be better off moving to a slower mining software with no fee or 100% accepted shares. Is there anyone that has this issue and has found a solution?

So was this the software just straight up stealing shares and reporting them as "lost because of job change" or is there a huge bug when the software switches to the donation pool to get its 2% cut? Either way this is an FYI to everyone be prepared to lose anywhere from 9% to 15% in shares because "Out of date job change" AND the dev fee of 2% this dev decided to include.

IMO fix your shares lost issue or remove the DEV fee.

I think stale shares are common with all miners/pools.  It's just that some miners (Claymore in particular) do a better job of hiding that information from the user.

You need to look at the hash rate reported by the pool.  If Cast were truly stealing shares you would see that 9% difference at the pool.  Better yet, try a pool that shows the actual hashes submitted and run a 24 hour test, that should tell you a lot.

My over all hash rate reported by the pool is lower then I would expect that's the reason I am looking into things more. I have tried MANY pools. Do you have an example of a better pool I can use for XMR or ETN?
newbie
Activity: 4
Merit: 0
What sort of HBM temperatures should one be aiming at to be able to run memory frequencies above 930-940 on unmodded Vega 56?
I'm mining without crashes as long as I don't push higher than 925 on my 6x Vega56 rig.

Currently set as follows:

GPU_P7=1407;900
Mem_P3=925;900
Power_Target=0

HBM temperature is between 70 and 80C. My cards are stuck pretty close together on that D1800 board.

Result: 1900 H/s per card, stable.

Is there anything I can do to improve the above? Lowering the power target just ends up starving core freq and doesn't help increasing mem speed.

Any help appreciated!

Get em cooler. I believe HBM throttles at 70C.
hero member
Activity: 1274
Merit: 556
What sort of HBM temperatures should one be aiming at to be able to run memory frequencies above 930-940 on unmodded Vega 56?
I'm mining without crashes as long as I don't push higher than 925 on my 6x Vega56 rig.

Currently set as follows:

GPU_P7=1407;900
Mem_P3=925;900
Power_Target=0

HBM temperature is between 70 and 80C. My cards are stuck pretty close together on that D1800 board.

Result: 1900 H/s per card, stable.

Is there anything I can do to improve the above? Lowering the power target just ends up starving core freq and doesn't help increasing mem speed.

Any help appreciated!
sr. member
Activity: 437
Merit: 250
any one tried miningrig rentals to direct miner to different pool?
my thinking was you direct your miner to mining rig rental, if pool fails mining rig rental automaticly connects with next pool on the list?
im testing it now
full member
Activity: 675
Merit: 100
So I have been able to get this running on my 11 4GB XFX RX 470s, I have one 6 card system and one 5 card system. On both I seem to get about 9% of my shares rejected because of "out of date job change". Is there no resolution to this? 9% is not a small amount either. This is well after 24 hours of straight mining. I get an 865 hash rate from each card 9% of that times 11 is over 870 H\s. In a sense I would be better off moving to a slower mining software with no fee or 100% accepted shares. Is there anyone that has this issue and has found a solution?

So was this the software just straight up stealing shares and reporting them as "lost because of job change" or is there a huge bug when the software switches to the donation pool to get its 2% cut? Either way this is an FYI to everyone be prepared to lose anywhere from 9% to 15% in shares because "Out of date job change" AND the dev fee of 2% this dev decided to include.

IMO fix your shares lost issue or remove the DEV fee.

I think stale shares are common with all miners/pools.  It's just that some miners (Claymore in particular) do a better job of hiding that information from the user.

You need to look at the hash rate reported by the pool.  If Cast were truly stealing shares you would see that 9% difference at the pool.  Better yet, try a pool that shows the actual hashes submitted and run a 24 hour test, that should tell you a lot.
full member
Activity: 1124
Merit: 136
So I have been able to get this running on my 11 4GB XFX RX 470s, I have one 6 card system and one 5 card system. On both I seem to get about 9% of my shares rejected because of "out of date job change". Is there no resolution to this? 9% is not a small amount either. This is well after 24 hours of straight mining. I get an 865 hash rate from each card 9% of that times 11 is over 870 H\s. In a sense I would be better off moving to a slower mining software with no fee or 100% accepted shares. Is there anyone that has this issue and has found a solution?

So was this the software just straight up stealing shares and reporting them as "lost because of job change" or is there a huge bug when the software switches to the donation pool to get its 2% cut? Either way this is an FYI to everyone be prepared to lose anywhere from 9% to 15% in shares because "Out of date job change" AND the dev fee of 2% this dev decided to include.

IMO fix your shares lost issue or remove the DEV fee.

Probably just your connection to where ever you are mining. Most i've seen is 5% loss.
newbie
Activity: 7
Merit: 0
So I have been able to get this running on my 11 4GB XFX RX 470s, I have one 6 card system and one 5 card system. On both I seem to get about 9% of my shares rejected because of "out of date job change". Is there no resolution to this? 9% is not a small amount either. This is well after 24 hours of straight mining. I get an 865 hash rate from each card 9% of that times 11 is over 870 H\s. In a sense I would be better off moving to a slower mining software with no fee or 100% accepted shares. Is there anyone that has this issue and has found a solution?

So was this the software just straight up stealing shares and reporting them as "lost because of job change" or is there a huge bug when the software switches to the donation pool to get its 2% cut? Either way this is an FYI to everyone be prepared to lose anywhere from 9% to 15% in shares because "Out of date job change" AND the dev fee of 2% this dev decided to include.

IMO fix your shares lost issue or remove the DEV fee.
hero member
Activity: 789
Merit: 501
Finally ! I just find some time to add this awesome miner to my Monitorig tool.
After some weeks without it, my Vega rig is not blind anymore Tongue

Here is my personal dashboard
http://www.monitorig.com/#?panel=dashboard&id=1086db9224632d26671ab42df5ee1d92bf0187bbdd

For power consumption reason, I'm limiting the hashrate, in 1800 (sometime drop to 1600 after some minutes, I'll investigate ... but I don't have dummy plug to let the screen on)  range per card (850W at wall for 5 cards)
I'll release and publish the new version of monitorig tomorrow after some testing tonight
legendary
Activity: 2294
Merit: 1182
Now the money is free, and so the people will be
worst case just stick a hdmi dummy plug if you have one laying around, maybe that can help you.  Seriously though, i've never had this bug that hashrate goes down.  rock solid for me, but temps are low and I am not too greedy with the hashrate vs power consumption
newbie
Activity: 99
Merit: 0
Anyone figure how to solve the hash rate drop to 1600?

HAhah man read the threadzzzzzzzzz.

Make sure the card is kept as cool as possible, under 70C, shoot for 65C and whatever fan speed works to achieve the needed temps to prevent the card from throttling memory speed. HEAT = BAD

Consider looking up the regedit to reduce power consumption cited many many times previously. Also, there are other tools to check the actual memory temps that are mentioned in the threads here. READ THREADZZZZ = HELPZZZ COMPUTERZZZZ HASHZZZZZ

Make sure you have disabled screen savers and make sure windows is set to never turn the monitor off. Never lock the computer or send it to the lock screen. Just turn off the monitor. SCREENLOCK = BAD

Then you can hash consistently without the long term drops. There are many steps just from here to fine tune the setup for XMR.
POWER HASHZZZ = MONeyEAZZZY

F Roll Eyes r crying out loud man!!!

That's the thing, they run at 59C, sometimes for hours, then randomly drop to 1600.

So look up the other tools mentioned in the previous 50 pages of posts that talk about other tools used to check on the actual temps occurring on the memory, which is different from Wattman/Overdriventool. You will need to check that obviously. Also as I mentioned check your windows settings to make sure monitor and HDD or SSD never turn off. Windows settings are just as important.
newbie
Activity: 14
Merit: 0
Anyone figure how to solve the hash rate drop to 1600?

HAhah man read the threadzzzzzzzzz.

Make sure the card is kept as cool as possible, under 70C, shoot for 65C and whatever fan speed works to achieve the needed temps to prevent the card from throttling memory speed. HEAT = BAD

Consider looking up the regedit to reduce power consumption cited many many times previously. Also, there are other tools to check the actual memory temps that are mentioned in the threads here. READ THREADZZZZ = HELPZZZ COMPUTERZZZZ HASHZZZZZ

Make sure you have disabled screen savers and make sure windows is set to never turn the monitor off. Never lock the computer or send it to the lock screen. Just turn off the monitor. SCREENLOCK = BAD

Then you can hash consistently without the long term drops. There are many steps just from here to fine tune the setup for XMR.
POWER HASHZZZ = MONeyEAZZZY

F Roll Eyes r crying out loud man!!!

That's the thing, they run at 59C, sometimes for hours, then randomly drop to 1600.
newbie
Activity: 99
Merit: 0
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
1 Vega64 on 17.11.4 1050H/s with Claymore 10.2 BETA. Cast XMR freeze my system for some reason...  I am trying to set it up for nanopool eu server, with not so much luck.

1050H/s for Vega 64? Are you saying that your hash rate before that was only ~800 H/s?

No Man. 1050 was with 17.11.4. Now with 17.12.1 its 1250

What are you even doing, you should be getting 1900 H/s with Vega 64.
With Blockchain driver. Im on the regular driver (17.12.1), beause im gaming on my computer too.

Anyhow, does anybody have any idea why CAST XMR freeze the computer?

Stop wasting 1's and 0's with posts like these man. Sad
newbie
Activity: 4
Merit: 0
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
1 Vega64 on 17.11.4 1050H/s with Claymore 10.2 BETA. Cast XMR freeze my system for some reason...  I am trying to set it up for nanopool eu server, with not so much luck.

1050H/s for Vega 64? Are you saying that your hash rate before that was only ~800 H/s?

No Man. 1050 was with 17.11.4. Now with 17.12.1 its 1250

What are you even doing, you should be getting 1900 H/s with Vega 64.
With Blockchain driver. Im on the regular driver (17.12.1), beause im gaming on my computer too.

Anyhow, does anybody have any idea why CAST XMR freeze the computer?
newbie
Activity: 99
Merit: 0
Anyone figure how to solve the hash rate drop to 1600?

HAhah man read the threadzzzzzzzzz.

Make sure the card is kept as cool as possible, under 70C, shoot for 65C and whatever fan speed works to achieve the needed temps to prevent the card from throttling memory speed. HEAT = BAD

Consider looking up the regedit to reduce power consumption cited many many times previously. Also, there are other tools to check the actual memory temps that are mentioned in the threads here. READ THREADZZZZ = HELPZZZ COMPUTERZZZZ HASHZZZZZ

Make sure you have disabled screen savers and make sure windows is set to never turn the monitor off. Never lock the computer or send it to the lock screen. Just turn off the monitor. SCREENLOCK = BAD

Then you can hash consistently without the long term drops. There are many steps just from here to fine tune the setup for XMR.
POWER HASHZZZ = MONeyEAZZZY

F Roll Eyes r crying out loud man!!!
newbie
Activity: 14
Merit: 0
Anyone figure how to solve the hash rate drop to 1600?
full member
Activity: 1124
Merit: 136
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
1 Vega64 on 17.11.4 1050H/s with Claymore 10.2 BETA. Cast XMR freeze my system for some reason...  I am trying to set it up for nanopool eu server, with not so much luck.

1050H/s for Vega 64? Are you saying that your hash rate before that was only ~800 H/s?

No Man. 1050 was with 17.11.4. Now with 17.12.1 its 1250

What are you even doing, you should be getting 1900 H/s with Vega 64.
newbie
Activity: 4
Merit: 0
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
1 Vega64 on 17.11.4 1050H/s with Claymore 10.2 BETA. Cast XMR freeze my system for some reason...  I am trying to set it up for nanopool eu server, with not so much luck.

1050H/s for Vega 64? Are you saying that your hash rate before that was only ~800 H/s?

No Man. 1050 was with 17.11.4. Now with 17.12.1 its 1250
newbie
Activity: 58
Merit: 0
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
1 Vega64 on 17.11.4 1050H/s with Claymore 10.2 BETA. Cast XMR freeze my system for some reason...  I am trying to set it up for nanopool eu server, with not so much luck.

1050H/s for Vega 64? Are you saying that your hash rate before that was only ~800 H/s?
newbie
Activity: 4
Merit: 0
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
1 Vega64 on 17.11.4 1050H/s with Claymore 10.2 BETA. Cast XMR freeze my system for some reason...  I am trying to set it up for nanopool eu server, with not so much luck.
newbie
Activity: 4
Merit: 0
With Claymore 10.2 Beta i have +200H/s on 17.12.1. Today is a happy day, eth  price is incrasing, plus hashrate and i didnt have to work today.Smiley

+200 compared to what?  Total hashrate?
Jump to: