Author

Topic: Claymore's CryptoNote AMD GPU Miner v11.3 - page 298. (Read 2145062 times)

full member
Activity: 182
Merit: 100
I received several msg asking me to give my workaround for connecting claymore to minergate.
I can give you the solution i found, but it's just for the proof of concept.

Claymore can't connect to the pool for devfee, so the miner try to reconnect every 5 sec, and after some amount of time, stop mining with a message.(something like 5 minutes or so)
Claymore can't merge mine neither, we will need a special miner for that (i think).

But mining on minergate with claymore would be already very good. The best thing you can do if you also want to, is to ask the dev if he could make a version which can.
Claymore is closed source, with 5% fee. I have no problems with that, but please listen to us !

Ok, I'll add this pool Smiley
Thank you VERY MUCH
donator
Activity: 1610
Merit: 1325
Miners developer
I received several msg asking me to give my workaround for connecting claymore to minergate.
I can give you the solution i found, but it's just for the proof of concept.

Claymore can't connect to the pool for devfee, so the miner try to reconnect every 5 sec, and after some amount of time, stop mining with a message.(something like 5 minutes or so)
Claymore can't merge mine neither, we will need a special miner for that (i think).

But mining on minergate with claymore would be already very good. The best thing you can do if you also want to, is to ask the dev if he could make a version which can.
Claymore is closed source, with 5% fee. I have no problems with that, but please listen to us !

Ok, I'll add this pool Smiley
full member
Activity: 182
Merit: 100
I received several msg asking me to give my workaround for connecting claymore to minergate.
I can give you the solution i found, but it's just for the proof of concept.

Claymore can't connect to the pool for devfee, so the miner try to reconnect every 5 sec, and after some amount of time, stop mining with a message.(something like 5 minutes or so)
Claymore can't merge mine neither, we will need a special miner for that (i think).

But mining on minergate with claymore would be already very good. The best thing you can do if you also want to, is to ask the dev if he could make a version which can.
Claymore is closed source, with 5% fee. I have no problems with that, but please listen to us !
legendary
Activity: 1151
Merit: 1001
yes.... merged mining with FCN is the sweet deal with MG
hero member
Activity: 812
Merit: 1000
Do you take in account that on extreme pool your shares "wait" 60 blocks to mature? afaik on minergate shares are immediately sent toy your account

For BCN, 9 hours later, total is ~70K for gpu and 57K for cpu on minergate. I should have more than 100K with GPU.

I managed to connect claymore to minergate, it works, but of course devfee can't connect to it.
Quote
DevFee: Failed to invoke login, retry in 5sec...
So, minergate see my miner, but less of what my miner can do (400H/s instead of 500), because the miner spend a lot of time trying to mine for the dev. - and 0 H/s for the dev obviously, and that not the goal.

Quote
06/18/14-11:05:51 - SHARE FOUND (target 1936) - (GPU 0)
SOMETHING IS WRONG WITH DEVFEE CONNECTION. STOP MINING.
and sometimes, a disconnect Wink
Quote
Speed: 528 h/s, TotalHashes: 455K, DevHashes: 0K Mining time: 00:14
Dev, for making a minergate version of your miner, you "just have" to make a version that send a login + mdp instead of a wallet address (i mean, for the dev fee) or mining for the devfee on a pool of your choice instead of mining on the same as user - for connecting to minergate, i found a workaround.
Please let us the choice where to mine, it's win/win
Can you explain how to connect to minergate?!

Doesn't work here...cpu runs well though. I get better hash with wolf's miner though, compared to their stock one.
sr. member
Activity: 364
Merit: 250
Any ideas if there is anyway to check my hash rate the last 24 hours ?!
legendary
Activity: 1151
Merit: 1001
Do you take in account that on extreme pool your shares "wait" 60 blocks to mature? afaik on minergate shares are immediately sent toy your account

For BCN, 9 hours later, total is ~70K for gpu and 57K for cpu on minergate. I should have more than 100K with GPU.

I managed to connect claymore to minergate, it works, but of course devfee can't connect to it.
Quote
DevFee: Failed to invoke login, retry in 5sec...
So, minergate see my miner, but less of what my miner can do (400H/s instead of 500), because the miner spend a lot of time trying to mine for the dev. - and 0 H/s for the dev obviously, and that not the goal.

Quote
06/18/14-11:05:51 - SHARE FOUND (target 1936) - (GPU 0)
SOMETHING IS WRONG WITH DEVFEE CONNECTION. STOP MINING.
and sometimes, a disconnect Wink
Quote
Speed: 528 h/s, TotalHashes: 455K, DevHashes: 0K Mining time: 00:14
Dev, for making a minergate version of your miner, you "just have" to make a version that send a login + mdp instead of a wallet address (i mean, for the dev fee) or mining for the devfee on a pool of your choice instead of mining on the same as user - for connecting to minergate, i found a workaround.
Please let us the choice where to mine, it's win/win
Can you explain how to connect to minergate?!
sr. member
Activity: 400
Merit: 263
Two pages back I posted the same problem with another rig.

For some odd reason some of my rigs show the wrong available memory for my 280x GPUs that have 3072 MB.

Below is the screenshot of a 5 card rig (3 x MSI 280x Gaming, 2 x Asus DC2 Top 280x). The rig has 8 MB RAM, pagefile set at 48.000 MB and has been running all other algos without a hitch in more than 7 months.
I wouldn't bother about the info in the screenshot below but each GPU only delivers about 150 h/s whereas my other 280x rigs do 320-350 H/s/GPU.

What am i supposed to do?

Side by side comparison of 2 rigs, all with 280x, above MSI and Asus, below 4 x Sapphire Vapor-x, rest of the hardware identical. All GPUs have 3073 MB VRAM.



donator
Activity: 1610
Merit: 1325
Miners developer
This error appears very often when rig have low RAM (2-4 Gb), even when size of pagefile is more then 16G (24 for example). Can you fix this?

This is not a problem of miner, it does not use a lot of RAM or page-file. It is problem of AMD OpenCL driver and I cannot find any workaround so far.
hero member
Activity: 616
Merit: 500
Running 3 270s with 1120/1500 averaging 730h/s for all three.

Edit: Just updated to 2.2 and that puts me at 820ish it seems. Nice Smiley

Edit: 830ish.
newbie
Activity: 18
Merit: 0
why i have just get 400H/s with 2*290 + 280X??
sr. member
Activity: 425
Merit: 250
Hi Guys!

First Claymore THANKS! I think you are doing a really great job both with this miner and the XPM one. (I really hope you make enough for the effort and time you put in!)

I have one test rig where this miner runs flawlessly.. (I can confirm it is working on 6870,6850 @ 70-100 H/s/card only 3 GB memory dual core.)

I wanted to try it on a "real" rig.

win7, catalyst 13.12 , 4x 79xx..  6 GB .. newest 2.2 version (got it also with 2.1)

06/17/14-02:11:28 - SHARE FOUND (target 5000) - (GPU 0)
GPU 0, GpuMiner k1a failed -4
GPU 0, GpuMiner k1a failed -4
GPU 0, GpuMiner k1a failed -4
GPU 0, GpuMiner k1a failed -4
GPU 0, GpuMiner k1a failed -4
06/17/14-02:11:28 - SHARE FOUND (target 5000) - (GPU 0)
Error in server response
: {"id":1,"jsonrpc":"2.0","error":{"code":-1,"message":"Lowdifficultyshare"}}

GPU 0, GpuMiner k1a failed -4

But I get this error I have not seen anyone else reporting...

Anyone with an idea what I could do?

@Clayore: you probably have the best chance of guessing or knowing what kould be wrong. Please if you have any "feeling" whatsoever.

Hope someone can help!

It's GPU memory allocation error, I will update miner so it will show more detailed message. Try to set Virtual Memory (pagefile) 16GB or larger as described in readme.

This error appears very often when rig have low RAM (2-4 Gb), even when size of pagefile is more then 16G (24 for example). Can you fix this?
full member
Activity: 182
Merit: 100
I continue with my testing :

BCN this time (no DDOS on pool so far)

CPU on minergate (with merge mining) : between 900H/s and 1200H/s seen by pool : 57K after 6 hours
GPU with pool (extrempool) : ~2400H/s : 46K after 6 hours.

I really don't understand why ther is such a difference.
I already tested with MRO on two different pool (cryptopool-fr and wolf), on QCN, and now with BCN. I always have less or barely more with GPU than with cpu on minergate..instead of 2 times more at least.


Next test : cpu and GPU on the same pool, with two differents address to control if the problem is pool or miner (i think it's pool, but i have to verify)


I asked dev for a version wich can mine to minergate's pool (even without merge mining - and still with a fee for him), but he doesn't want.
Solo is not possible neither.


How are you getting 1200H/s on Minergate with CPU?  I have an i7 4771 and I'm getting around 140 H/s
For CPU, i have a 4 node server with 4 x dual Xeon L5530 + dual Xeon 2687w + [email protected] = total 1000~1200H/s

Another test :
8 hours with Ducknote :

CPU on minergate : with only the dual-xeon 2687w + [email protected] : between 500 to 800H/s ==> 290K ducknote
GPU (6 R9 280X + 1x5970= on extrempool : ~2400H/s ==> 130K ducknote

for the same hashrate, cpu on minergate make 6x more coin Huh? (and some FCN with merge mining as bonus)

Do you take in account that on extreme pool your shares "wait" 60 blocks to mature? afaik on minergate shares are immediately sent toy your account

For BCN, 9 hours later, total is ~70K for gpu and 57K for cpu on minergate. I should have more than 100K with GPU.

I managed to connect claymore to minergate, it works, but of course devfee can't connect to it.
Quote
DevFee: Failed to invoke login, retry in 5sec...
So, minergate see my miner, but less of what my miner can do (400H/s instead of 500), because the miner spend a lot of time trying to mine for the dev. - and 0 H/s for the dev obviously, and that not the goal.

Quote
06/18/14-11:05:51 - SHARE FOUND (target 1936) - (GPU 0)
SOMETHING IS WRONG WITH DEVFEE CONNECTION. STOP MINING.
and sometimes, a disconnect Wink
Quote
Speed: 528 h/s, TotalHashes: 455K, DevHashes: 0K Mining time: 00:14
Dev, for making a minergate version of your miner, you "just have" to make a version that send a login + mdp instead of a wallet address (i mean, for the dev fee) or mining for the devfee on a pool of your choice instead of mining on the same as user - for connecting to minergate, i found a workaround.
Please let us the choice where to mine, it's win/win
legendary
Activity: 1151
Merit: 1001
I continue with my testing :

BCN this time (no DDOS on pool so far)

CPU on minergate (with merge mining) : between 900H/s and 1200H/s seen by pool : 57K after 6 hours
GPU with pool (extrempool) : ~2400H/s : 46K after 6 hours.

I really don't understand why ther is such a difference.
I already tested with MRO on two different pool (cryptopool-fr and wolf), on QCN, and now with BCN. I always have less or barely more with GPU than with cpu on minergate..instead of 2 times more at least.


Next test : cpu and GPU on the same pool, with two differents address to control if the problem is pool or miner (i think it's pool, but i have to verify)


I asked dev for a version wich can mine to minergate's pool (even without merge mining - and still with a fee for him), but he doesn't want.
Solo is not possible neither.


How are you getting 1200H/s on Minergate with CPU?  I have an i7 4771 and I'm getting around 140 H/s
For CPU, i have a 4 node server with 4 x dual Xeon L5530 + dual Xeon 2687w + [email protected] = total 1000~1200H/s

Another test :
8 hours with Ducknote :

CPU on minergate : with only the dual-xeon 2687w + [email protected] : between 500 to 800H/s ==> 290K ducknote
GPU (6 R9 280X + 1x5970= on extrempool : ~2400H/s ==> 130K ducknote

for the same hashrate, cpu on minergate make 6x more coin Huh? (and some FCN with merge mining as bonus)

Do you take in account that on extreme pool your shares "wait" 60 blocks to mature? afaik on minergate shares are immediately sent toy your account
full member
Activity: 182
Merit: 100
I continue with my testing :

BCN this time (no DDOS on pool so far)

CPU on minergate (with merge mining) : between 900H/s and 1200H/s seen by pool : 57K after 6 hours
GPU with pool (extrempool) : ~2400H/s : 46K after 6 hours.

I really don't understand why ther is such a difference.
I already tested with MRO on two different pool (cryptopool-fr and wolf), on QCN, and now with BCN. I always have less or barely more with GPU than with cpu on minergate..instead of 2 times more at least.


Next test : cpu and GPU on the same pool, with two differents address to control if the problem is pool or miner (i think it's pool, but i have to verify)


I asked dev for a version wich can mine to minergate's pool (even without merge mining - and still with a fee for him), but he doesn't want.
Solo is not possible neither.


How are you getting 1200H/s on Minergate with CPU?  I have an i7 4771 and I'm getting around 140 H/s
For CPU, i have a 4 node server with 4 x dual Xeon L5530 + dual Xeon 2687w + [email protected] = total 1000~1200H/s

Another test :
8 hours with Ducknote :

CPU on minergate : with only the dual-xeon 2687w + [email protected] : between 500 to 800H/s ==> 290K ducknote
GPU (6 R9 280X + 1x5970= on extrempool : ~2400H/s ==> 130K ducknote

for the same hashrate, cpu on minergate make 6x more coin Huh? (and some FCN with merge mining as bonus)
sr. member
Activity: 364
Merit: 250
What setting are ppl getting with Gigabytes 280x-ox-3DG

i cant run  stock setting 1100/1500 @ 1200
1030/1500 @1200 is stable but getting less then 330h/s  per card

Asus 280x dc2 get 376h/s stock  but these Gigabytes are a pain in the ass
Elpida ram?
Probably it's because of the Elpida. A lot of people here complained about big differences in performance between cards with one thing in common so far, the ram type.
It seems that Elpida is vastly under performing.
legendary
Activity: 882
Merit: 1000
What setting are ppl getting with Gigabytes 280x-ox-3DG

i cant run  stock setting 1100/1500 @ 1200
1030/1500 @1200 is stable but getting less then 330h/s  per card

Asus 280x dc2 get 376h/s stock  but these Gigabytes are a pain in the ass
Elpida ram?

exact same cards as you, good luck getting the gigabytes up past 320-330, seems to be their limit due to poor ram quality. and asus DC2 can do 400h/s easy. the gb ones also have a habit of dropping out more then most as i have found....
my msi 7950's at stock clocks are faster then gb 280x lol.... go figure
hero member
Activity: 599
Merit: 500
What setting are ppl getting with Gigabytes 280x-ox-3DG

i cant run  stock setting 1100/1500 @ 1200
1030/1500 @1200 is stable but getting less then 330h/s  per card

Asus 280x dc2 get 376h/s stock  but these Gigabytes are a pain in the ass
Elpida ram?
sr. member
Activity: 364
Merit: 250
Oh wow, diff went sky high... At 500mil right now Shocked was like 3.3mil yesterday (at least it was at some point).
legendary
Activity: 1151
Merit: 1001
from dashboard. Withdrawals have problems every 3 days or so... but so far they were resolved
Jump to: