Pages:
Author

Topic: ### A ChainWorks Industries (CWI) Project - CWIgm | Simple Powerful Stable - page 26. (Read 67732 times)

newbie
Activity: 71
Merit: 0
Currently testing intensity 26 on 1080's and 1080ti's, seems to squeeze a bit more hash

25.5 seems optimal for 1070's, 26 causes instability
hero member
Activity: 714
Merit: 512
I am interested in trying this miner.

I currently have ~2 GH of power across 24 computers (1060s, 1070s, 1080s, and one 1080 Ti) on SIGT on the ccminer.org pool using the krnlx miner.
full member
Activity: 227
Merit: 100
x6 rig still running strong... lost track of how many hours.  timer on the miner says 31+ hours connected, but that's connected, not uptime.  that number was reset when the stratum was restarted yesterday.  the rig has been on much longer than that.

----------------------------------------------------------------------------------------
CPU:0.47%          ChainWorks Industries beta - Version 0.9.8         [13:43:46]
--------------------------------------------------------------------------------
Algo:Skunk      | Port: 2000 | NetDiff: 25775.4            | Block:      36587 |
--------------------------------------------------------------------------------
Coin:signatum   | POW: 1250  | Estimated coins/day: 208.527                    |
--------------------------------------------------------------------------------
Solved:   0 | Diff:     1.24 | Shares: 7201.34 | Rej: 0.476656% | Speed:  175.89
--------------------------------------------------------------------------------
0.EVGA GTX 1070              : Shares: 1179.31 | Rej: 8.63 | Speed:   29.17MH/s|
1.EVGA GTX 1070              : Shares: 1216    | Rej: 6.34 | Speed:   29.25MH/s|
2.EVGA GTX 1070              : Shares: 1246.42 | Rej: 7.48 | Speed:   29.38MH/s|
3.EVGA GTX 1070              : Shares: 1133.75 | Rej: 3.06 | Speed:   28.99MH/s|
4.EVGA GTX 1070              : Shares: 1183.82 | Rej: 5.31 | Speed:   29.52MH/s|
5.EVGA GTX 1070              : Shares: 1242.04 | Rej: 3.67 | Speed:   29.58MH/s|
--------------------------------------------------------------------------------
0.EVGA GTX 1070              : T:56C | F:59% | P:108W | I: 24.0 | Core:1911Mhz |
1.EVGA GTX 1070              : T:57C | F:60% | P:106W | I: 24.0 | Core:1910Mhz |
2.EVGA GTX 1070              : T:50C | F:50% | P:106W | I: 24.0 | Core:1915Mhz |
3.EVGA GTX 1070              : T:56C | F:59% | P:107W | I: 24.0 | Core:1892Mhz |
4.EVGA GTX 1070              : T:52C | F:53% | P:106W | I: 24.0 | Core:1926Mhz |
5.EVGA GTX 1070              : T:53C | F:55% | P:107W | I: 24.0 | Core:1928Mhz |
--------------------------------------------------------------------------------
[13:42:49]Mean netDiff: 20588.7        | [05:43:25][GPU1]:Job not found (=stale
[13:42:50]Share: 731073ab              | )
[13:43:09]Mean netDiff: 20924          | [06:18:32][GPU5]:Invalid nonce
[13:43:14]Share: e506d387              | [06:48:45][GPU4]:Job not found (=stale
[13:43:16]Share: c5c65036              | )
[13:43:18]Share: f8a636ba              | [10:35:23][GPU0]:Invalid nonce
[13:43:28]Share: 689d1a76              | [10:42:12][GPU2]:Invalid nonce
[13:43:39]Mean netDiff: 21210.4        | [12:18:14][GPU4]:Invalid nonce
--------------------------------------------------------------------------------

everything looks good Smiley

yesterday i pulled my smaller rigs (2 x1's and 1 x2 for a total of 4) out and pointed them at suprnova using the kmlx miner.  

in the last 24 hours my 4 gpu's pointed there earned me 110 and change coins.  suprnova reports 92.5 percent luck for the day.

in the same 24 hours my 6xgpu rig pointed here earned 224 and change coins.  pool reports luck of 107.5 percent.

so, i earned 110/4=27.5 coins per gpu on suprnova, and 224/6=37.3 coins per gpu here.

now, this is by faaaaar not a "scientific" test in any way.  the x6 gpu rig is inherently much more efficient then an armada of small rigs, and i would've probably done better with it per gpu on suprnova vs smaller rigs.  so again - not science here, just an observation using the equipment i happen to have on hand.

as  "unscientific" as they are, these results are encouraging.  there is a big difference percentage wise between 27.5 and 37.3.  and this is despite the fact that suprnova had almost a 14 percent advantage in luck (92.5 vs 107.5).  

my second x6gpu rig (an identical clone of the first) should be up and running by end of the week, so it would be interesting to do this again with the "twin" rigs and have all variables be equal.  but for now, especially considering that this is a beta miner running on a pool that could really use some tweaking, the results are pretty impressive Smiley

again, not a scientific test in any way, so draw your own conclusions...


full member
Activity: 280
Merit: 100
Try without --lodiff


Did not help, it's still not submitting any shares.

full member
Activity: 182
Merit: 100
It's possible to use CWIgm miner to mine on other CWI pools or it's currently locked to SIGT and DNR?
full member
Activity: 728
Merit: 106
any chance you will add XCN  to your pool/miner?  it is a great nvidia coin that has been around forever. 
There are a lot of other algos that are used in many perspective coins like skein or nist5. Which will attract many miners to CWI if they could make faster miner for these algos.
For what do you ask about this almost dead algo used in ONE coin with 0.5btc volume on some shitty exchange and constantly low price for years?
sr. member
Activity: 326
Merit: 250
I will write this just. It might be a small bug i seen.

When i started miner several times and after restart also, it always initialized only 2x GPU from 7. 7 were shown in device manager and in after burner, but when i started miner only gpu 2 and 6 shown hashrate, rest were stuck on 0.


Then i cancled miner for 5th time, went alt tab to do something, tried run miner again and it worked. Dunno why it didnt recognized my GPU-s right away.


This is my .bat file:


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
cwigm -c sigt -u Brucelats.Zotacx7_SIGT -p x --cpu-priority=3 -i 22 --tlimit=65,65,65,65,65,65,65
pause


I used exact same on my other rigs with 5 6 7 GPU-s, only changed temp limits a bit and worker name and they all worked. Only thing that is speciffic to this rig that had problems is, it had all the same GPU-s and that are Zotac GTX 1070 AMP! Extreme. Rest have mix, or at least 1 card is different. Cheers!

first - remove all the setx environment parameters ...

they are really only used for sgminer / amd settings ...

cuda settings are usually set within the environment when you install the nvidia driver - and there is no need to set those environment parameter for nvidia cards ... you will see the same hashrate and results when you remove them ...

second - why --tlimit? ...

CWIgm is a unique miner ... try without that parameter and restart ...

you may want to try adding --lodiff to access the lower difficulty port also ... at least test it -until the port changes take place ( if they take place ) ...

let us know that all goes ...

#crysx


Ok will change parameters. And i forgot --lodiff parameter. I'll add it yeah.



On second note, why not use temp limit?

I cant monitor my rig all the time. In case something happens and gpus start overheating i want my miner to close to prevent any damage. I use Air conditioner coz it's quite hot. In case something occurs with AC temps will start rising, so i want use temp limits to stop miners if that happens.

Unless i am wrong?

If there is no parameter to turn off miner if certain temperature limit is reached it should be added. I would always use it, and everyone should in case something goes wrong. Ok there are parameters to controll GPU usage and such, but i have my fans set on auto, i made a curve. And if i reach some high temperatures, something bad is going on haha. So no point to mine until that thing is resolved, so best idea is just to turn off miner and no harm done.

there is a device called the PID ... it is a controller of sorts - an algorithm ...

https://en.wikipedia.org/wiki/PID_controller#Fundamental_operation ...

the implemented PID controller - at this point - has a racing condition with the internal firmware which controls the fans speed ... which means that if the user has not set the fan speed manually at a desired level - the pid will drop the temperature to the limit set by the user - but the gpu will notice that the temperature is going down and will drop the fan speeds - which will result in a temperature increase which the PID has to readjust and so on ...

ie ... do not use the --max-temp feature with the fans set to auto ...

the temp control and the netdiff parameters are experiential ... the netdiff parameter calculates the netdiff incorrectly - even tho its only about 1% error - we advise NOT using it at all for now ... it is fixed in CWIgm-0.9.9 ...

other than that - great results ...

soon - you will only need the --lodiff parameter for small rigs ... the consensus is overwhelmingly in favour of stratum diff change - so there will be some teething issues there - but we will get through them Smiley ...

#crysx


Thank you for the answer, but i havent got answer what i was really looking for? What does --tlimit does? Will it turn off the miner or something if certain temeprature is met?

I have great settings with my auto fan speed in MSI afterburner and all my cards are around 50-57 degrees. I am looking for function that will TURN OFF MY MINER if temeperature reaches certain point. I dont want any throtling or extra control of fans etc. I JUST want to turn off my miner if something messes up and to save my gpu-s or whatever.

IS --tlimit doing that? If not YOU should REALLY implement that function. It should be easy. Just turn off miner if he reads temerature limit is being reached or passed.


Cheers!
full member
Activity: 227
Merit: 100
Is it only me or are there no rewards for the last one+ hour? diff :317.44 and a constant 120+ MH/s but no coins?  Shocked Shocked

pool had a bad run of luck... took like an hour and 50 minutes to find the last block.

this will happen.  usually (and hopefully) rarely, but it will happen once in a while in any pool, especially a small pool.

small pool = large variance.



newbie
Activity: 31
Merit: 0
Thanks for adding me to the closed beta all is up and running now.     My speeds are roughly the same as using ccminer 1.1 but they started off higher.    My 8 gpu rig has some cooling issues on a few of the GPU that I need to address and the miner is keeping them at 70C which is hurting their performance.   It started off at about a 8% improvement and has slowly reduced back down but is keeping the GPUs much cooler.    This will force me to actually address the air flow issue.  

How often are the coin added into the unconfirmed?   I have been mining exactly 60min and have not had anything show up there.    I mined for a while yesterday and the same thing, although I logged in today and there was confirmed coin there so it does appear that its being paid out but their must be a long delay before it shows up.  

I love the interface and layout of the miner

Thanks!
full member
Activity: 621
Merit: 108
Is it only me or are there no rewards for the last one+ hour? diff :317.44 and a constant 120+ MH/s but no coins?  Shocked Shocked

No, not only you. I just accidentaly noted the same thing. What's the matter, admins?
full member
Activity: 142
Merit: 100
Is it only me or are there no rewards for the last one+ hour? diff :317.44 and a constant 120+ MH/s but no coins?  Shocked Shocked
newbie
Activity: 9
Merit: 0
may i ask a question?

http://imgur.com/nqvoz8N

what is the "Job not found" and "Invalid nonce" means?

and my bat is

cwigm -c sigt -u pp688039.pp688039 -p x --lodiff --max-temp=70
pause


thanks

why OC your mem, is there any advantage?what i know SIGT is core extensive...

i think  the mem can more or less increase the performance (maybe)

maybe is wrong...just i think...

I'm negative on my mem and up core to OC and get better rates doing that

Example

-180 mem
+215 core


ok, i try to low down the mem to test it

core +140
mem -502

the mem have to low down to limit?
or a little is enough?
full member
Activity: 148
Merit: 100
how to join the close beta? i already sent a PM with no reply Sad sad
full member
Activity: 462
Merit: 100
Parachute for sale. Used once. Small red stain.
may i ask a question?

http://imgur.com/nqvoz8N

what is the "Job not found" and "Invalid nonce" means?

and my bat is

cwigm -c sigt -u pp688039.pp688039 -p x --lodiff --max-temp=70
pause


thanks

why OC your mem, is there any advantage?what i know SIGT is core extensive...

i think  the mem can more or less increase the performance (maybe)

maybe is wrong...just i think...

I'm negative on my mem and up core to OC and get better rates doing that

Example

-180 mem
+215 core
full member
Activity: 462
Merit: 100
Parachute for sale. Used once. Small red stain.
Didn't read through entire thread but I'm mining sigt and love testing new miners. Anyway for me to get added to the beta?

Use GeForce GTX1070 8GB cards, win 10. Current drivers for nvidia.
newbie
Activity: 9
Merit: 0
may i ask a question?

http://imgur.com/nqvoz8N

what is the "Job not found" and "Invalid nonce" means?

and my bat is

cwigm -c sigt -u pp688039.pp688039 -p x --lodiff --max-temp=70
pause


thanks

why OC your mem, is there any advantage?what i know SIGT is core extensive...

i think  the mem can more or less increase the performance (maybe)

maybe is wrong...just i think...

full member
Activity: 434
Merit: 107
member
Activity: 153
Merit: 11
hi Dev,
Recently, I have checked out your 0.9.8. that is awesome, so I try to pull more SIGT miner recently back to the CWI pool, so the speed is getting faster. I would like to have a feedback/request.
- Whenever I go back to the pool website, I need to type all email/pass again. Other pool allow to remember, and the process to login is much more simple. Can you solve it
- Waiting to hear your 0.9.9 release Smiley
 thanks

At my side it remembers the login cause I'm logged with my google account on chrome ^^
newbie
Activity: 48
Merit: 0
hi Dev,
Recently, I have checked out your 0.9.8. that is awesome, so I try to pull more SIGT miner recently back to the CWI pool, so the speed is getting faster. I would like to have a feedback/request.
- Whenever I go back to the pool website, I need to type all email/pass again. Other pool allow to remember, and the process to login is much more simple. Can you solve it
- Waiting to hear your 0.9.9 release Smiley
 thanks
newbie
Activity: 25
Merit: 0
may i ask a question?

http://imgur.com/nqvoz8N

what is the "Job not found" and "Invalid nonce" means?

and my bat is

cwigm -c sigt -u pp688039.pp688039 -p x --lodiff --max-temp=70
pause


thanks

why OC your mem, is there any advantage?what i know SIGT is core extensive...
Pages:
Jump to: