Topic: Gateless Gate Sharp 1.3.8: 30Mh/s (Ethash) on RX 480! - page 142. (Read 214431 times)

zawawa,

you are sure for the conf with devfee ?
i run many hour and no devfee !
now i have read and change my conf with load-balance at true.

https://github.com/sgminer-dev/sgminer/blob/master/README.md

The thing about SA's Python implementation is that it does not work well at all on Windows.
In fact, the miner got faster by 10% just by switching from SA to sgminer-gm while keeping the same Equihash kernel.
I am not entirely satisfied with sgminer as I said before, and I will probably write my own backend and user-interface in Clojure in a long run.

No problem.

Looking pretty good so far. Thanks a lot!

Can i change to asia pool for the donation part? coz i'm from asia side..

Going to do a 24 hr test with my full farm here shortly, Just finalizing the configs. Should throw about 30MH/s as dev fee.
Edit: Sending the command out now to switch to GG

All the pool stability issues aside, I think the technical side of the new fee system is working.
The low numbers of valid shares and active computers are as expected.                                   .
I just need more people to mine with the fee on.
At the current pace, I would be making an impressive $3/month. LOL

https://ethermine.org/miners/91fa32e00b0f365d629fb625182a83fed61f0642/payouts

Oh, the fee realized through bat-files... I thought that the fee is hardcoded into miners code... I'm sorry for stupid questions about disabling the fee. Now I can disable it in any moment. But i proceed some time mining with fee. You work must be rewarded.
Thanks for your hard work.

P.S. Please add nicehash compatibility.

Xi dont have a huge impact on mining, use 256 , you get same hashrate and less hardware errors

Try --gpu-threads 2 --xintensity 4608 . I got equivalent rates to Claymore with that setting. 4608 is a multiple of the amount of stream processors on an rx480. Not sure what 4620 could be a multiple of.

unstable system with a lot of MemHWs - yes I know it  
0,00 HW-rate of claymore vs 100,000 HW-rate of zawawa - which one should I prefer?

Aways use dual thread if you want to get the maximum of it, even claymore miner uses dual thread on Gpu

I really have to emphasize somewhere that software/bios configurations for Claymore's DO NO TRANSFER to GG.
If you want to get the most out of GG, you have to customize the timing mode not for Claymore's but for GG.
Simple, no?

Gigabyte RX480 (Hynix) custom timing-mode @1100/1970
claymore 9.0 - 29.05mh/s -asm 1 -dcri 6
zawawa pre6 - 27.60mh/s -Ix 4620 -g 1 +more power/+more system impact&freezes/there are memory HW (due to hwinfo output)
there is still a huge room for optimizations  

I too noticed connection pb with ethermine and devFee.

@Zawawa: there must be something not working well with devfee, because if you look at your adress on ethermine you can clearly see some people are/try mining but there's to few share's.

Not sure if this is common, but you may see less donated shares because of it:
[08:33:05] Stratum connection to ethermine.org interrupted
[08:33:06] Rejected untracked stratum share from ethermine.org
[08:33:09] Stratum connection to ethermine.org interrupted
[08:33:10] ethermine.org not responding!
[08:33:10] Switching to ethermine.org (donation)
[08:33:12] ethermine.org alive, testing stability
[08:34:20] ethermine.org stable for 60 seconds
[08:34:20] Switching to ethermine.org

I also think it would be best to move away from sgminer. It's great that it supports a ton of currencies, but you're going to end up burdened supporting this massive codebase that comes with a lot of quirks.

What mbevand did by moving all the sensitive code to C and using Python for communication was brilliant imo. It's simple, readable, and testable.

One last thing is that the Core folder may be more neat organizationally, but it also adds another level of complexity for the less technologically inclined (to put it nicely) users. Perhaps a separate project for each optimized coin, like Claymore's miners, with all the guts in the root folder would do you well.

Hello zawawa,

I've been following this thread since your early ZEC mining goals, hoping that one day the titan that is Claymore would be beat (or at least matched considering his ZEC history) by an open source implementation. All of these bottom feeders asking for how to disable a voluntary mining fee without even looking finally got me to register and contribute.

Benchmarked gateless-0.1.3-pre6 against Claymore v8.1 and v9.0 and took power readings.

My system with 3 XFX RX480 runs at 117W idle. One of the cards won the bad ASIC quality lottery and runs at a much lower rate.

These numbers are at the wall usage on an 80 plus platinum efficiency EVGA 750P2 power supply:

483W mining v8.1
ETH - Total Speed: 86.514 Mh/s, Total Shares: 1, Rejected: 0, Time: 00:00
ETH: GPU0 29.697 Mh/s, GPU1 27.157 Mh/s, GPU2 29.659 Mh/s

483W mining v9.0
ETH - Total Speed: 86.282 Mh/s, Total Shares: 1, Rejected: 0, Time: 00:00
ETH: GPU0 29.612 Mh/s, GPU1 27.052 Mh/s, GPU2 29.618 Mh/s

504W mining gateless-0.1.3-pre6 1 thread xI 2304
[06:05:25] GPU0                | (5s):29.40M (avg):27.30Mh/s | A:4000000000 R:0 HW:2 WU:30.089/m
[06:05:25] GPU1                | (5s):27.11M (avg):25.27Mh/s | A:4000000000 R:0 HW:0 WU:23.524/m
[06:05:25] GPU2                | (5s):29.42M (avg):27.30Mh/s | A:8000000000 R:0 HW:3 WU:22.977/m

509W mining gateless-0.1.3-pre6 2 thread xI 4608
[06:13:33] GPU0                | (5s):29.69M (avg):26.50Mh/s | A:0 R:0 HW:7 WU:26.056/m
[06:13:33] GPU1                | (5s):27.32M (avg):24.91Mh/s | A:12000000000 R:0 HW:5 WU:21.564/m
[06:13:33] GPU2                | (5s):29.69M (avg):26.87Mh/s | A:0 R:0 HW:9 WU:23.360/m

Edit:
After 1hr 20mins:
GPU 0:                | 29.57M/29.46Mh/s | R:  0.0% HW:165 WU:27.588/m xI:3204
GPU 1:                | 27.23M/27.15Mh/s | R:  0.0% HW:116 WU:24.575/m xI:3204
GPU 2:                | 29.54M/29.43Mh/s | R:  0.0% HW:125 WU:27.935/m xI:3204

About 5 HW errors every minute across 3 cards on xI 3204 with 1 gpu thread. They're all undervolted with custom straps so it could be my problem.

I'll keep it mining with you as 1% so you can hopefully keep moving forward. At only a few cents loss every 24hr compared to Claymore it's worth it to support a competitor who isn't afraid to release their code!

Your miner consumes more electricity! with each video card 10w

It may be the case that the most recent optimization may not work for RX 470 due to timing issues.
I am out of country right now, but I brought an RX 470 with me.
Let me see what I can do with it.

I really don't know what to say, but thanks for considering me for the position. I mean it.