Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 563.

Vann

hero member

Activity: 1036

Merit: 606

Quote from: lmlim on October 19, 2017, 08:12:58 PM

Quote from: allknowingeye on October 19, 2017, 01:23:55 PM

Windows 1709 update that came out today breaks Claymore 10. It says the opencl.dll is now missing and will not start.

Update: Reinstalling the video card driver corrected the problem. Turns out the Windows 1709 Update dumps the video driver.

is amd blockchain driver still compatible with win10 ver 1709? yesterday i reformatted 1 rig because of this unwantted upgrade.

The AMD blockchain driver works with the v1709 build but the problem is the Windows UI becomes unresponsive and unstable. I did a clean install back to v1703 until a new driver is released. Supposedly there will be an updated driver later this year that will increase the maximum limit of 8 of the same GPU type in Windows 10.

lmlim

newbie

Activity: 16

Merit: 0

Quote from: allknowingeye on October 19, 2017, 01:23:55 PM

Windows 1709 update that came out today breaks Claymore 10. It says the opencl.dll is now missing and will not start.

Update: Reinstalling the video card driver corrected the problem. Turns out the Windows 1709 Update dumps the video driver.

is amd blockchain driver still compatible with win10 ver 1709? yesterday i reformatted 1 rig because of this unwantted upgrade.

allknowingeye

sr. member

Activity: 361

Merit: 250

Windows 1709 update that came out today breaks Claymore 10. It says the opencl.dll is now missing and will not start.

Update: Reinstalling the video card driver corrected the problem. Turns out the Windows 1709 Update dumps the video driver.

greatauror28

full member

Activity: 181

Merit: 101

Ethereum Miner

Quote from: Ursul0 on October 19, 2017, 07:54:21 AM

Quote from: bardacuda on October 19, 2017, 07:03:18 AM

Quote from: Zorr0 on October 18, 2017, 05:50:34 PM

Quote from: greatauror28 on October 03, 2017, 12:53:08 PM

Since building my first rig last June, i've always had this rock-stable settings for six of my Asus Dual OC 580s 8GB Samsung memory

-cclock 1260
-mclock 2050
-cvddc 925mV
-mvddc 925mV

This yields me a constant 29.1 to 29.3 across all cards (175 total) and pulls about 1175 up to 1200w. High, I know but they go for weeks until needing a restart.

Earlier this week, I played with them again to squeeze in few more Mh. These are my settings:
-cclock 1140
-mclock 2075
-cvddc 875
-mvddc 875

They average a total of 177.5 now and cards hashes from 29.5 up to 30.0. But the best thing is they now only pull 1035w dual-mining with DCR.

Wow thats excellent results for those cards. I've checked some forums and noone was able to get those numbers.
I've tried different tweaked BIOS (even from RX480) which gave same results but nothing was stable for more than few hrs.
With blockchain driver the only stable bios was original with 1750 straps and -cvddc 1000 (28MHz).
Which BIOS do you use?

That's actually not great. Power usage could be much lower. For example I have a 4-card rig that pulls ~580W from the wall (with a bronze PSU so would prob be more like 550W with a gold). The rig pulls 80W @ idle so each card pulls about 125W under load. Stable for over 530 hours w/ over 29MH/s per card. If this were a 6-card rig it should only be pulling between 780W (gold PSU) to 830W (bronze PSU).

https://i.imgur.com/pO6WSdX.jpg

EDIT: Ahh nvm...just noticed he is dual mining DCR. Different story.

same cards giving me pain when installed 8 per rig, leftover 4 run smoothly at over 30MH. usual 1500 straps, ~850-880mv, 1180&2100 (elpida is nice these days)

EDIT: Just as I was saying a year ago: RX *70 anytime over RX *80

When I still had my 8-gpu rig consisting of MSI Gaming X 580 4gb elpidas, I'm having similar problems as you. They're smooth and all but they only hash 29.6 if all eight cards are connected.

If I only plug 4-5 cards, they hash a terrific 30.1 up to 30.3 each.

I've created a thread here, in another board and even in ethermining sub of reddit but apparently, not too many miners have 8-gpu rigs to contribute their thoughts.

I'm in the process of building 8x Nitro+ 580s on Saturday and I hope and pray that it's not going to be the same experience as the Gaming X's.

BossmanPL

newbie

Activity: 37

Merit: 0

Quote from: haxiboy on October 18, 2017, 04:58:33 PM

Quote from: BossmanPL on October 18, 2017, 04:14:10 PM

Claymore i need to monitor rigs with php, is there any example for such language, i know there is nodejs monitoring tool but wasn't able to find any php example, do you have something or can provide as an test/example ?, thanks.

It's just Json, you can write monitoring in any language that can create json queries and parse the result. Its so Easy

Edit : PHP Can do it, HTML5 can do it, and almost every programming language have abilities to interact with json.

i did not know that, json is easy, but how to get to that json, if i set remote port, how do i connect to claymore api, socket way ?
is there a way to switch off or on card or restart claymore via api ?

Ursul0

sr. member

Activity: 857

Merit: 262

Quote from: bardacuda on October 19, 2017, 07:03:18 AM

Quote from: Zorr0 on October 18, 2017, 05:50:34 PM

Quote from: greatauror28 on October 03, 2017, 12:53:08 PM

Since building my first rig last June, i've always had this rock-stable settings for six of my Asus Dual OC 580s 8GB Samsung memory

-cclock 1260
-mclock 2050
-cvddc 925mV
-mvddc 925mV

This yields me a constant 29.1 to 29.3 across all cards (175 total) and pulls about 1175 up to 1200w. High, I know but they go for weeks until needing a restart.

Earlier this week, I played with them again to squeeze in few more Mh. These are my settings:
-cclock 1140
-mclock 2075
-cvddc 875
-mvddc 875

They average a total of 177.5 now and cards hashes from 29.5 up to 30.0. But the best thing is they now only pull 1035w dual-mining with DCR.

Wow thats excellent results for those cards. I've checked some forums and noone was able to get those numbers.
I've tried different tweaked BIOS (even from RX480) which gave same results but nothing was stable for more than few hrs.
With blockchain driver the only stable bios was original with 1750 straps and -cvddc 1000 (28MHz).
Which BIOS do you use?

That's actually not great. Power usage could be much lower. For example I have a 4-card rig that pulls ~580W from the wall (with a bronze PSU so would prob be more like 550W with a gold). The rig pulls 80W @ idle so each card pulls about 125W under load. Stable for over 530 hours w/ over 29MH/s per card. If this were a 6-card rig it should only be pulling between 780W (gold PSU) to 830W (bronze PSU).

https://i.imgur.com/pO6WSdX.jpg

EDIT: Ahh nvm...just noticed he is dual mining DCR. Different story.

same cards giving me pain when installed 8 per rig, leftover 4 run smoothly at over 30MH. usual 1500 straps, ~850-880mv, 1180&2100 (elpida is nice these days)

EDIT: Just as I was saying a year ago: RX *70 anytime over RX *80

bardacuda

sr. member

Activity: 430

Merit: 254

Quote from: Zorr0 on October 18, 2017, 05:50:34 PM

Quote from: greatauror28 on October 03, 2017, 12:53:08 PM

Since building my first rig last June, i've always had this rock-stable settings for six of my Asus Dual OC 580s 8GB Samsung memory

-cclock 1260
-mclock 2050
-cvddc 925mV
-mvddc 925mV

This yields me a constant 29.1 to 29.3 across all cards (175 total) and pulls about 1175 up to 1200w. High, I know but they go for weeks until needing a restart.

Earlier this week, I played with them again to squeeze in few more Mh. These are my settings:
-cclock 1140
-mclock 2075
-cvddc 875
-mvddc 875

They average a total of 177.5 now and cards hashes from 29.5 up to 30.0. But the best thing is they now only pull 1035w dual-mining with DCR.

Wow thats excellent results for those cards. I've checked some forums and noone was able to get those numbers.
I've tried different tweaked BIOS (even from RX480) which gave same results but nothing was stable for more than few hrs.
With blockchain driver the only stable bios was original with 1750 straps and -cvddc 1000 (28MHz).
Which BIOS do you use?

That's actually not great. Power usage could be much lower. For example I have a 4-card rig that pulls ~580W from the wall (with a bronze PSU so would prob be more like 550W with a gold). The rig pulls 80W @ idle so each card pulls about 125W under load. Stable for over 530 hours w/ over 29MH/s per card. If this were a 6-card rig it should only be pulling between 780W (gold PSU) to 830W (bronze PSU).

https://i.imgur.com/pO6WSdX.jpg

EDIT: Ahh nvm...just noticed he is dual mining DCR. Different story.

P00P135

full member

Activity: 1125

Merit: 136

Quote from: Victorio on October 19, 2017, 04:02:34 AM

Quote from: Call_Me_Bambi on October 18, 2017, 05:54:51 PM

In my mix of cards I've only one Gigabyte cards, an RX480, and it has not been melting, but I've two XFX RX580's and they all have the melting thermal pad problem.
Some people have had this problem with Sapphire RX480.

I have this problem with Sapphire RX480 Nitro (plain Nitro, not +). Almost all these cards are dripping oil, to a different extent. I use to clean them with ethanol every 2-3 months, including the metallic base of the heatsink - from my experience, oil gathers around heatpipes and makes a nasty grease in combination with dust.
I also have XFX RX480 RS. Although XFX cooling system is much poorer than Sapphire Nitro (memory chips are not connected to heatsink and may heat up to 110-120 C), it does not have the problem of melting pads.
Needed to say that I keep my Sapphires at 63-65C and the XFX at 57C (core temp).

Did someone use both ethanol and isopropyl to clean this silicone oil? Which alcohol works best?

I remember another post someone claimed it was oil from the fans leaking out. Probably best to RMA them.

Victorio

full member

Activity: 281

Merit: 100

BTW, regarding hashrate drops:

From parsing Claymore miner logs, I notice that RX480 happen to drop from 28-29 MH/s to 20 MH/s about 0.2-1.0% of all hashrate line reports. That is a 30% decrease in reported hashrate, that occurs with a frequency of ~0.5% all the time. Since RX460 and RX560 show the same, I consider this to be a specific issue of AMD Polaris architecture and don't pay it attention, as long as this drops account for <1% of all mining time.
If the drop is permanent, then it is a real issue - agree.

I haven't noticed such drops on Nvidia GTX1050 Ti and GTX1060 with Samsung memory.

Victorio

full member

Activity: 281

Merit: 100

Quote from: Call_Me_Bambi on October 18, 2017, 05:54:51 PM

In my mix of cards I've only one Gigabyte cards, an RX480, and it has not been melting, but I've two XFX RX580's and they all have the melting thermal pad problem.
Some people have had this problem with Sapphire RX480.

I have this problem with Sapphire RX480 Nitro (plain Nitro, not +). Almost all these cards are dripping oil, to a different extent. I use to clean them with ethanol every 2-3 months, including the metallic base of the heatsink - from my experience, oil gathers around heatpipes and makes a nasty grease in combination with dust.
I also have XFX RX480 RS. Although XFX cooling system is much poorer than Sapphire Nitro (memory chips are not connected to heatsink and may heat up to 110-120 C), it does not have the problem of melting pads.
Needed to say that I keep my Sapphires at 63-65C and the XFX at 57C (core temp).

Did someone use both ethanol and isopropyl to clean this silicone oil? Which alcohol works best?

Zorr0

newbie

Activity: 10

Merit: 0

Quote from: kkourmousis on October 11, 2017, 12:07:17 PM

Hey mate,

Since you are reading for such a long time you must have read my post - chain, where (in my final post) I concluded that the blockchain drivers can cause power spikes that cause great instability.

Mr. Claymore said on this:

Quote from: Claymore on September 28, 2017, 10:07:06 AM

Quote from: kkourmousis on September 28, 2017, 07:51:10 AM

As I said, I am not happy to see that those problems are getting noticed by more users everyday. For myself, lowering the voltages seems to did the trick but I still experience some freezes now and then (I guess that I didn't lower them enough).
I am hoping that with so many cases around, Mr. Claymore can verify the existence of the problem.
What I would like to hear from Mr. Claymore is if there are any thoughts for a solution and what I would love to hear is an ETA if one exists.

1. I still cannot reproduce this issue, I described what tests I did: https://bitcointalksearch.org/topic/m.21966972
If you have ste-by-step guide how to reproduce this issue - let me know.
2. I'm not sure if it is possible to fix this issue at all, since all these issues are from buggy ADL.

So there you have it.

Reducing voltage indeed does the trick on system freezes.
I myself had to lower my clocks as well with the voltage, sacrificing some hashrate in order to have stability

Thank you very much for your reply. I did miss your "final post" which totally makes sense. (It should have its own topic about power spikes)
I have applied your settings and got much better stability right away, but still not stable for 24hr straight Sad

Making -cvddc 1000 -mvddc 1000 and/or reducing core frequency does the trick, but sacrifices the hashrate and increases power, which doesn't make sense to me after reading your "final post".
I have a feeling blockchain drivers are not as stable as regular ones, but we have no choice here.

zeeff

newbie

Activity: 2

Merit: 0

Quote from: manna4all on October 18, 2017, 01:25:52 AM

Hi guys,

Have any of you encountered reduced hash rate? I just noticed now that my rig went from ~29Mh/s per card to 24~25. But when mining ubiq, I still get 29mh/s.

Yes. Same here. 3 mh lower.

Call_Me_Bambi

sr. member

Activity: 574

Merit: 261

Quote from: Victorio on October 17, 2017, 03:33:49 PM

Quote from: rzrwolf on October 17, 2017, 01:56:57 PM

it was shitty Gigabyte thermal interfaces. Never buy Gigabyte.

Did someone else encounter the problem is melting thermal pads on Gigabyte cards? Gigabyte RX570/RX580 Aorus in particular.

In my mix of cards I've only one Gigabyte cards, an RX480, and it has not been melting, but I've two XFX RX580's and they all have the melting thermal pad problem.
Some people have had this problem with Sapphire RX480.
Pic of my 'oily' RX580 in this thread (post #1095). https://bitcointalk.org/index.php?topic=1874892.1080

Zorr0

newbie

Activity: 10

Merit: 0

Quote from: greatauror28 on October 03, 2017, 12:53:08 PM

Since building my first rig last June, i've always had this rock-stable settings for six of my Asus Dual OC 580s 8GB Samsung memory

-cclock 1260
-mclock 2050
-cvddc 925mV
-mvddc 925mV

This yields me a constant 29.1 to 29.3 across all cards (175 total) and pulls about 1175 up to 1200w. High, I know but they go for weeks until needing a restart.

Earlier this week, I played with them again to squeeze in few more Mh. These are my settings:
-cclock 1140
-mclock 2075
-cvddc 875
-mvddc 875

They average a total of 177.5 now and cards hashes from 29.5 up to 30.0. But the best thing is they now only pull 1035w dual-mining with DCR.

Wow thats excellent results for those cards. I've checked some forums and noone was able to get those numbers.
I've tried different tweaked BIOS (even from RX480) which gave same results but nothing was stable for more than few hrs.
With blockchain driver the only stable bios was original with 1750 straps and -cvddc 1000 (28MHz).
Which BIOS do you use?

haxiboy

newbie

Activity: 36

Merit: 0

Quote from: BossmanPL on October 18, 2017, 04:14:10 PM

Claymore i need to monitor rigs with php, is there any example for such language, i know there is nodejs monitoring tool but wasn't able to find any php example, do you have something or can provide as an test/example ?, thanks.

It's just Json, you can write monitoring in any language that can create json queries and parse the result. Its so Easy

Edit : PHP Can do it, HTML5 can do it, and almost every programming language have abilities to interact with json.

carlo_0000

legendary

Activity: 1281

Merit: 1003

Quote from: mano237 on October 17, 2017, 03:23:44 PM

I use some 970 and get only 3-4mh. I know its a common bug on windows 10 and newer drivers..

Is it still not possible on windows 10 or is there already a fix?

for nvidia you need to update windows 10 to last version

BossmanPL

newbie

Activity: 37

Merit: 0

Claymore i need to monitor rigs with php, is there any example for such language, i know there is nodejs monitoring tool but wasn't able to find any php example, do you have something or can provide as an test/example ?, thanks.

mamoEB

newbie

Activity: 10

Merit: 0

Claymore any news about updates or new version?

blackhwk77

newbie

Activity: 27

Merit: 0

So now that it has soaked a couple days, I am definitely seeing a payout increase since the fork. For the most part it seems equivalent to rolling back to early Sept before the Sept 22 difficulty increase.

fr4nkthetank

legendary

Activity: 2294

Merit: 1182

Now the money is free, and so the people will be

Quote from: asdpof on October 18, 2017, 05:34:00 AM

One of my cards keeps having a moving hash rate. It goes from 14.5 Mh/s to 30.2 Mh/s every 5 seconds and then goes back. Always and repetedly. It is a pretty constant oscilation, always with the same range and about the same frequency.

Does anyone experiencing something similar?

you probably have an -etha setting, get rid of it. or something like that. remove all settings to just the bare minimum. other than that, check your GPU temps, are you hitting thermal limits and throttling ? underclock and undervolt your GPU. good luck

Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 563. (Read 6590757 times)