Pages:
Author

Topic: Genesis Mining Presents: SGMiner-GM - now with Zawawa's GG! [Updated 17/01/2017] - page 35. (Read 140431 times)

legendary
Activity: 980
Merit: 1001
aka "whocares"
I'm getting steady 30MH/s on each Fury X on stock clocks (1050/500). Claymore's can only get about 28MH.
A 390 at 1040Mhz gets 29Mh, with some overclocking I can surely can 32-33MH, but my cards warm up too much. Definitely the best miner out there.

The only issue I can spot is the CPU usage, which makes this miner basically worthless on weak CPUs. I couldn't run it on a single-core AMD CPU and a friend of mine had his rig crashing with a dual core Celeron installed. most mining rigs have very weak CPUs, and if this miner gets updated to support them, I am sure it will get much more traction.

There is my problem, all of my rigs are on Celeron processors.  I thought I was doing something wrong when it was not my settings at all.  I got the 31 on a test bench with an i5 but cannot duplicate it on any rig
legendary
Activity: 1797
Merit: 1028
I'm getting steady 30MH/s on each Fury X on stock clocks (1050/500). Claymore's can only get about 28MH.
A 390 at 1040Mhz gets 29Mh, with some overclocking I can surely can 32-33MH, but my cards warm up too much. Definitely the best miner out there.

The only issue I can spot is the CPU usage, which makes this miner basically worthless on weak CPUs. I couldn't run it on a single-core AMD CPU and a friend of mine had his rig crashing with a dual core Celeron installed. most mining rigs have very weak CPUs, and if this miner gets updated to support them, I am sure it will get much more traction.

I can fix it with a new binary - just needs to be recompiled against a better pthreads lib.

WILL THE SGMINER-GM RECEIVE UPDATES? --

I run this SGminer on a sensitive 280X rig that is not stable with Claymore and dual mining.    Currently, SGminer-gm is running under Ethos 1.1.1 on 4 280X cards, mining at 16.7MH/s at stock clocks.  It runs with less heat than either Genoil v1.1.7 or Claymore, and does not crash as often.  The CPU is a Haswell Celeron 1820 at 2.7GHz.  The rig is not perfectly stable, but it runs.

Are updated binaries gong to be posted for both Linux and Windows?  I am hoping that the repository will be an active one.       --scryptr
YIz
hero member
Activity: 686
Merit: 502
I'm getting steady 30MH/s on each Fury X on stock clocks (1050/500). Claymore's can only get about 28MH.
A 390 at 1040Mhz gets 29Mh, with some overclocking I can surely can 32-33MH, but my cards warm up too much. Definitely the best miner out there.

The only issue I can spot is the CPU usage, which makes this miner basically worthless on weak CPUs. I couldn't run it on a single-core AMD CPU and a friend of mine had his rig crashing with a dual core Celeron installed. most mining rigs have very weak CPUs, and if this miner gets updated to support them, I am sure it will get much more traction.

I can fix it with a new binary - just needs to be recompiled against a better pthreads lib.

I'd love to see this miner fixed. would be adopted among miners very quickly.
member
Activity: 81
Merit: 1002
It was only the wind.
Why is the miner abandoned? I'm getting higher hashrates than Claymore's and it's open source! A really good piece of software right there, apart for CPU usage.

I told him to use SSE in the regenhash, but NOOOO... it should work on *ALL* the CPUs...
That's the wrong conclusion. The high CPU usage only happens on Windows, so that one core shows 100% usage. Regenhash is executed every ~2s with one GPU.
It seems to be a bug in libwinpthread because earlier sgminer had this issue on Windows, too.

Oh, my builds actually don't do that. Ever. It is pthread.
YIz
hero member
Activity: 686
Merit: 502
I'm getting steady 30MH/s on each Fury X on stock clocks (1050/500). Claymore's can only get about 28MH.
A 390 at 1040Mhz gets 29Mh, with some overclocking I can surely can 32-33MH, but my cards warm up too much. Definitely the best miner out there.

The only issue I can spot is the CPU usage, which makes this miner basically worthless on weak CPUs. I couldn't run it on a single-core AMD CPU and a friend of mine had his rig crashing with a dual core Celeron installed. most mining rigs have very weak CPUs, and if this miner gets updated to support them, I am sure it will get much more traction.
legendary
Activity: 980
Merit: 1001
aka "whocares"
Up to 31.2 on a 390x running 1100/1250 - definitely making headway.  A little more tuning and we have a winner Grin
hero member
Activity: 578
Merit: 508
Even with no hash improvement, lot's of operational info in one place.

I take it that the ability to use the GPU-vddc command, will be card dependent?
legendary
Activity: 980
Merit: 1001
aka "whocares"
I get similar/marginally higher speeds using stock settings using 1 RX470 + 1 7950. Also when moving the kernel into my fork. It's good!

Need to try that out, except yours doesn't have API support to use with AwesomeMiner - I wish it did Wink
member
Activity: 81
Merit: 1002
It was only the wind.
Why is the miner abandoned? I'm getting higher hashrates than Claymore's and it's open source! A really good piece of software right there, apart for CPU usage.

I told him to use SSE in the regenhash, but NOOOO... it should work on *ALL* the CPUs...
sr. member
Activity: 438
Merit: 250
I get similar/marginally higher speeds using stock settings using 1 RX470 + 1 7950. Also when moving the kernel into my fork. It's good!
legendary
Activity: 980
Merit: 1001
aka "whocares"
I am still playing with settings on a 390 and 390x and have yet to find any settings that come close to Genoil or Claymore's.  I am using 15.12 and have tried a full range of xintensity and worksizes and the best I have done on running it for an hour per setting is about 26mh.  So far definitely not worth converting the farm to it yet but I am sure it will be when I find the correct settings.

Usually when I have problems setting up something it's because I have overly complicated the task--
newbie
Activity: 49
Merit: 0
Why is the miner abandoned? I'm getting higher hashrates than Claymore's and it's open source! A really good piece of software right there, apart for CPU usage.

I told him to use SSE in the regenhash, but NOOOO... it should work on *ALL* the CPUs...
That's the wrong conclusion. The high CPU usage only happens on Windows, so that one core shows 100% usage. Regenhash is executed every ~2s with one GPU.
It seems to be a bug in libwinpthread because earlier sgminer had this issue on Windows, too.
newbie
Activity: 50
Merit: 0
@NaN_PTS

Thanks, and tomorrow I will check this.

@YIz

What's your card and settings?
YIz
hero member
Activity: 686
Merit: 502
Why is the miner abandoned? I'm getting higher hashrates than Claymore's and it's open source! A really good piece of software right there, apart for CPU usage.
sr. member
Activity: 588
Merit: 251
The WU (work units) are used to calculate the number of HW errors. Let's assume your card mines with an average of 20 MH/s and has an error rate of 2%. Then the expected WU value would be 20/m * (1-0.02) = 19.6/m. The hardware errors are normal, but they aren't shown by other Ethereum miners. A defective card typically produces 10 times more hardware errors then a working one.

Genoil's ethminer does report hardware errors.

Code:
FAILURE: GPU gave incorrect result!
newbie
Activity: 49
Merit: 0
@OhGodAGirl

Big thanks Smiley

My last questions:
1) sgminer shows hashrate like this: 21.70M/20.21Mh/s - is this 20.21Mh/s my card's hashrate? I got 22 Mh/s @ CM ...
2) WU: 20-21/m for such hashrate - is it OK?

TIA
The first value is an exponential moving average of your hash rate. Therefore, this value is a good measure of your current hash rate after a few minutes. The second value is an average over the whole run time and it should be equal to the first value after a few hour, if the pool connection is stable. One Radeon RX 480 8GB does about 23.8 MH/s with stock clocks and your card seems to be close to 21.7 MH/s (you have to run it longer).
The WU (work units) are used to calculate the number of HW errors. Let's assume your card mines with an average of 20 MH/s and has an error rate of 2%. Then the expected WU value would be 20/m * (1-0.02) = 19.6/m. The hardware errors are normal, but they aren't shown by other Ethereum miners. A defective card typically produces 10 times more hardware errors then a working one.
newbie
Activity: 50
Merit: 0
@OhGodAGirl

Big thanks Smiley

My last questions:
1) sgminer shows hashrate like this: 21.70M/20.21Mh/s - is this 20.21Mh/s my card's hashrate? I got 22 Mh/s @ CM ...
2) WU: 20-21/m for such hashrate - is it OK?

TIA
full member
Activity: 199
Merit: 108
Look, I'm really not that interesting. Promise.
I would start at 'gpu-threads: 1' and 'xintensity: 4620' because I'm not sure whether sgminer accepts rawintensity much larger than 2,000,000.

Wow, thanks! Now we are onto something Cheesy

With the below settings:

Code:
        "worksize": "192",
        "name": "eth",
        "algorithm": "ethash",
        "gpu-threads": "1",
        "xintensity": "4620"
    }],
    "no-extranonce": "true",
    "default-profile": "eth"

I got this:



Hashrates are close to Claymores - but they could be a little better Wink

Just those HW errors Sad - Any ideas?



HW's are normal, as long as they're trickling in and not skyrocketing, it's all good.
newbie
Activity: 50
Merit: 0
I would start at 'gpu-threads: 1' and 'xintensity: 4620' because I'm not sure whether sgminer accepts rawintensity much larger than 2,000,000.

Wow, thanks! Now we are onto something Cheesy

With the below settings:

Code:
        "worksize": "192",
        "name": "eth",
        "algorithm": "ethash",
        "gpu-threads": "1",
        "xintensity": "4620"
    }],
    "no-extranonce": "true",
    "default-profile": "eth"

I got this:



Hashrates are close to Claymores - but they could be a little better Wink

Just those HW errors Sad - Any ideas?

Pages:
Jump to: