Claymore's ZCash/BTG AMD GPU Miner v12.6 (Windows/Linux)

[Qumiu]

On Sapphire RX 470 4GB:
Claymore v12.1 get between 262-286 H/s, unstable
Claymore v12.2 get between 285-290 H/s, more stable
Claymore v12.3 get between 290-295 H/s, more stable
Claymore v12.4 get between 291-299 H/s, more stable

[1714623173]

1714623173

[1714623173]

1714623173

[1714623173]

1714623173

[1714623173]

1714623173

[jddebug]

Claymore,

I have 42 Radeon 7970's. I've been mining with your miners for a year. I get 1.17KHs with 12.0 on a 4 card rig but every version after that lowers my hashrate.  I just tried 12.4 and same problem. All cards hash slower. Getting 1.05 max.

Any suggestions? I have increased the voltage but that doesn't seem to do it.

Thanks.

Have you installed newer drivers? I get less h/s with those 17.x.x drivers than with 16.9.2

Per the readme I am using 15.12 on all rigs.

[djolep07]

for jddebug

What is your intesity setting i am using sapphire r9 280x and MSI 7970 lighting and have found that if I go for intesity higher than -i 2 my rigs start going unstable (restarting,hanging etc..) for the 7970 with downvolt to 1069 mV on the GPU core, on the r9 280 x I can do -i 4 and be stable with the downvolt of 980 mV.

7970 downvolt stable -i 2 get 290 to 300 h/s on claymore 12.4
r9 280 downvolt stable -i 4 get 290 to 300 h/s on claymore 12.4 

All cards are running asm option
Hope this helps

[Spacemarko]

Which driver is best for the new Claympre  v12.4 and 470 4GB Samsung base memory?

[kisk]

To all the people having issues with the 12.x versions... I was having issue with this as well. I was getting lock-ups, bluescreens, etc. Try bumping up your virtual memory to 25gb and make sure you have plenty of disk space to allocate it.

[Palider]

To all the people having issues with the 12.x versions... I was having issue with this as well. I was getting lock-ups, bluescreens, etc. Try bumping up your virtual memory to 25gb and make sure you have plenty of disk space to allocate it.

Does the big virtual memory solve the problem?

[orbital_station1]

Anyone using modded bios on R9 390? Hynix

[VL2D]

Hello to all 
Give me pls best moded rom for r9 280x
Thnx))

[coinzoid]

Dear Claymore,

v12.4 seems stable after running it more than 15 hours with my 6 gpu setup. However, I'd like to use values like 880, 890 for cvddc.

Besides to save more power with Sapphire RX 480 Nitro+ OC 8GB (Samsung Memory) [11260-01-20g] can someone share their values for cvddc and mvddc?

It looks like cvddc 900 works well but it didn't detect lower values on my test. Anyone else can use cvddc with values lower than 900?

My ultimate goal is keeping power cost low.


BTW, I noticed that cards with lower asics value has better downvolt results.

[Ursul0]

@Claymore
Miner Manager issue: Can you please move button "Save all settings" to the top of the page, otherwise it prevents proper narrow resize of the whole application window, as it stays inaccessible below the window frame...

Also I noticed some weird behavior of 12.4 on 370 - it produces much lower hashrate (110-140?)... so I'm back to 12.2 and it makes 162MH stable at 875:1375 stock.
While at the same time my Polaris cards enjoying the ~1% boost from 12.4

EDIT: also on Pitcairn 11.1 works more stable in general, also recovers on itself from significant rate drops due to simultaneous heavy GPU usage - like video playback, that can take it from 150-170 to 30-70 (particularly MPEG-2,  probably due to it's size) and then it will go back... in an hour or so ... very smooth... Started using v12 on my desktop and noticed that the rate doesn't come back. (disabling and re enabling device brings it all back, yet...) now back to 11.1, so I guess I'll see...

[Xardas2014]

with latest release whats the performance on Fiji(Fury/Nano) at default clock speed?

Do you own any Fiji(Fury/Nano)?

[citronick]

with latest release whats the performance on Fiji(Fury/Nano) at default clock speed? i am looking to compare with a RX/R9 3xx

R9-Nano
Default clock 1000/500
430-440 sols/s
Beast of a card - excellent for ZEC mining.

R9-Fury X.... supersonic at ~450-460 sols/s
RX480 ... at best 300 sols/s (RX cards best for ETH mining)
R9-390... very good ~380-390 sols/s

[Ursul0]

BTW, I noticed that cards with lower asics value has better downvolt results.

Lol it looks like that. I don't trust this "asic quality" bs anyways. I mean the asics ARE different, but whatever gpu-z shows has no meaning/use for me.
Depending on the algo (starting with ethash and getting the best out of cards) you can run core and mem from 854mv

[jddebug]

for jddebug

What is your intesity setting i am using sapphire r9 280x and MSI 7970 lighting and have found that if I go for intesity higher than -i 2 my rigs start going unstable (restarting,hanging etc..) for the 7970 with downvolt to 1069 mV on the GPU core, on the r9 280 x I can do -i 4 and be stable with the downvolt of 980 mV.

7970 downvolt stable -i 2 get 290 to 300 h/s on claymore 12.4
r9 280 downvolt stable -i 4 get 290 to 300 h/s on claymore 12.4 

All cards are running asm option
Hope this helps

This did help get me where I wanted to be. I was able to use -i 6 for my 7970's and they are stable so far.

Thank you.

[citronick]

with latest release whats the performance on Fiji(Fury/Nano) at default clock speed? i am looking to compare with a RX/R9 3xx

R9-Nano
Default clock 1000/500
430-440 sols/s
Beast of a card - excellent for ZEC mining.

R9-Fury X.... supersonic at ~450-460 sols/s
RX480 ... at best 300 sols/s (RX cards best for ETH mining)
R9-390... very good ~380-390 sols/s

Thanks for the information
then you didnt get an increase for the Fiji GPUs? i guess you maxed the power limit on the R9 Nano? I wonder why R9 Nano doesnt reach same performance a Fury X if both share similar setup

Nano vs Fury X -- big difference.
Both are fast cards for ZEC because of the HBM memory.
Fury X is the big brother. Still the fastest in the R9 class cards and it comes with its own cooling system.
Nano is small card but power packed and consumes less watts. Hard to get this card used - expensive too.
If you are using Linux - IMHO, best to go with Optiminer v1.7; otherwise Claymore v12.4 for Windows based rigs.

[thesaltonsea]

for jddebug

What is your intesity setting i am using sapphire r9 280x and MSI 7970 lighting and have found that if I go for intesity higher than -i 2 my rigs start going unstable (restarting,hanging etc..) for the 7970 with downvolt to 1069 mV on the GPU core, on the r9 280 x I can do -i 4 and be stable with the downvolt of 980 mV.

7970 downvolt stable -i 2 get 290 to 300 h/s on claymore 12.4
r9 280 downvolt stable -i 4 get 290 to 300 h/s on claymore 12.4 

All cards are running asm option
Hope this helps

This did help get me where I wanted to be. I was able to use -i 6 for my 7970's and they are stable so far.

Thank you.

Hi,
so you undervolt them and now they are stabe with more then -i 2 ?

[jddebug]

for jddebug

What is your intesity setting i am using sapphire r9 280x and MSI 7970 lighting and have found that if I go for intesity higher than -i 2 my rigs start going unstable (restarting,hanging etc..) for the 7970 with downvolt to 1069 mV on the GPU core, on the r9 280 x I can do -i 4 and be stable with the downvolt of 980 mV.

7970 downvolt stable -i 2 get 290 to 300 h/s on claymore 12.4
r9 280 downvolt stable -i 4 get 290 to 300 h/s on claymore 12.4 

All cards are running asm option
Hope this helps

This did help get me where I wanted to be. I was able to use -i 6 for my 7970's and they are stable so far.

Thank you.

Hi,
so you undervolt them and now they are stabe with more then -i 2 ?

No, I made changes that djolep07 made me think about.

I have voltage set to about the same as djolep07.

I have this in my config:

-zpool equihash.usa.nicehash.com:3357
-zwal (my wallet address).r10
-zpsw z
-allpools 1
-i 6
-asm 1

(I had -i 8. Even though the miner sets intensity to 7 that is to high and it made everything slower. I believe that was the problem.)

I didn't have -asm 1 in my config. I think the miner was automatically using it anyway but now I set it to use it for sure.

Additionally, I noticed that my start.bat file didn't have the extra parameters in it anymore. I may have missed that when I did a previous update.

So now it has all this in it:

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
ZecMiner64.exe

Getting a nice 300+ H/s from each card now.

[thesaltonsea]

for jddebug

What is your intesity setting i am using sapphire r9 280x and MSI 7970 lighting and have found that if I go for intesity higher than -i 2 my rigs start going unstable (restarting,hanging etc..) for the 7970 with downvolt to 1069 mV on the GPU core, on the r9 280 x I can do -i 4 and be stable with the downvolt of 980 mV.

7970 downvolt stable -i 2 get 290 to 300 h/s on claymore 12.4
r9 280 downvolt stable -i 4 get 290 to 300 h/s on claymore 12.4 

All cards are running asm option
Hope this helps

This did help get me where I wanted to be. I was able to use -i 6 for my 7970's and they are stable so far.

Thank you.

Hi,
so you undervolt them and now they are stabe with more then -i 2 ?

No, I made changes that djolep07 made me think about.

I have voltage set to about the same as djolep07.

I have this in my config:

-zpool equihash.usa.nicehash.com:3357
-zwal (my wallet address).r10
-zpsw z
-allpools 1
-i 6
-asm 1

(I had -i 8. Even though the miner sets intensity to 7 that is to high and it made everything slower. I believe that was the problem.)

I didn't have -asm 1 in my config. I think the miner was automatically using it anyway but now I set it to use it for sure.

Additionally, I noticed that my start.bat file didn't have the extra parameters in it anymore. I may have missed that when I did a previous update.

So now it has all this in it:

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
ZecMiner64.exe

Getting a nice 300+ H/s from each card now.

So you are using more than -i 2 on 7970 cards and it is not crashing ?

[FFI2013]

with latest release whats the performance on Fiji(Fury/Nano) at default clock speed? i am looking to compare with a RX/R9 3xx

R9-Nano
Default clock 1000/500
430-440 sols/s
Beast of a card - excellent for ZEC mining.

R9-Fury X.... supersonic at ~450-460 sols/s
RX480 ... at best 300 sols/s (RX cards best for ETH mining)
R9-390... very good ~380-390 sols/s

Thanks for the information
then you didnt get an increase for the Fiji GPUs? i guess you maxed the power limit on the R9 Nano? I wonder why R9 Nano doesnt reach same performance a Fury X if both share similar setup

Nano vs Fury X -- big difference.
Both are fast cards for ZEC because of the HBM memory.
Fury X is the big brother. Still the fastest in the R9 class cards and it comes with its own cooling system.
Nano is small card but power packed and consumes less watts. Hard to get this card used - expensive too.
If you are using Linux - IMHO, best to go with Optiminer v1.7; otherwise Claymore v12.4 for Windows based rigs.

Which card would you recommend the nano or fury Im looking to retire my 270s and add something to my 390/290 also are you running you 390 at stock clocks it seems to get mine at 400 I pushed it to 1150

[prichina]

BTW, I noticed that cards with lower asics value has better downvolt results.

Lol it looks like that. I don't trust this "asic quality" bs anyways. I mean the asics ARE different, but whatever gpu-z shows has no meaning/use for me.
Depending on the algo (starting with ethash and getting the best out of cards) you can run core and mem from 854mv

BTW, I noticed that cards with lower asics value has better downvolt results.

Cards with lower Asics have VRM's that signals higher voltage then cards with higher asics. Cards with higher asics their VRM signals lower voltage, so lowering the voltage do tend to make problems. So that is why, cards with lower asics works better with undervolting Cant' explain better then this in layman's terms

- But, all summed up, the total power consumption is equivalent in both cases, wheter you think that the card with lower asic can be undervolt lower....basically you get the same result as the cards with higher asics, ...in the end, the power consumption is the same.