Author

Topic: [ANN]Bminer: a fast Equihash/Ethash/Cuckaroo29z miner for AMD/NVIDIA GPUs 16.4.9 - page 141. (Read 148347 times)

member
Activity: 461
Merit: 49
I would strongly advise people to look at their mining pool hashrate for this miner and not the miners reported hashrate... Do the test...    Shocked

I'm beginning to think the same thing but it's more than that, when you first run it you get really nice looking numbers, then after a couple of hours, what you say sets in.

Here is the comparison of DSTM vs BMiner.

First line is DSTM that had been running for a few days
Next line is Bminer after 3 hours.  I saw grossly deviating current hash as well.

This was a GTX 950 that I just happened to have, but the two sets of number are on the same computer with the same hardware, etc.

Code:
Current Hashrate    Average Hashrate
      0H/s              155.9H/s
      80H/s             147.9H/s

As a final test, I'm going to leave it running in -nofee mode while I do some other work on a different miner

Can you run your numbers again with devfee turned on?

The miner will disable a number of optimizations if you disable devfee.

Maybe I mis spoke - those numbers above are *with devfee turned ON* and are what FlyPool is reporting, not the miner software.  Probably the biggest problem is that the radically changing instant hash rate, I see it cycle up and down by huge margins like over 100h/s in this case and this is just a small GPU.

I'm running in -nofee mode right now just to see what happens

The mining pool is estimating your hashrate based on the number of shares that the miner has submitted.

Given that you have only one GPU, it will take a while to stabilize. I would suggest running both miners for at least 24 hours to get the results.
member
Activity: 140
Merit: 18
I would strongly advise people to look at their mining pool hashrate for this miner and not the miners reported hashrate... Do the test...    Shocked

I'm beginning to think the same thing but it's more than that, when you first run it you get really nice looking numbers, then after a couple of hours, what you say sets in.

Here is the comparison of DSTM vs BMiner.

First line is DSTM that had been running for a few days
Next line is Bminer after 3 hours.  I saw grossly deviating current hash as well.

This was a GTX 950 that I just happened to have, but the two sets of number are on the same computer with the same hardware, etc.

Code:
Current Hashrate    Average Hashrate
      0H/s              155.9H/s
      80H/s             147.9H/s

As a final test, I'm going to leave it running in -nofee mode while I do some other work on a different miner

Can you run your numbers again with devfee turned on?

The miner will disable a number of optimizations if you disable devfee.

Maybe I mis spoke - those numbers above are *with devfee turned ON* and are what FlyPool is reporting, not the miner software.  Probably the biggest problem is that the radically changing instant hash rate, I see it cycle up and down by huge margins like over 100h/s in this case and this is just a small GPU.

I'm running in -nofee mode right now just to see what happens
member
Activity: 461
Merit: 49
I would strongly advise people to look at their mining pool hashrate for this miner and not the miners reported hashrate... Do the test...    Shocked

I'm beginning to think the same thing but it's more than that, when you first run it you get really nice looking numbers, then after a couple of hours, what you say sets in.

Here is the comparison of DSTM vs BMiner.

First line is DSTM that had been running for a few days
Next line is Bminer after 3 hours.  I saw grossly deviating current hash as well.

This was a GTX 950 that I just happened to have, but the two sets of number are on the same computer with the same hardware, etc.

Code:
Current Hashrate    Average Hashrate
      0H/s              155.9H/s
      80H/s             147.9H/s

As a final test, I'm going to leave it running in -nofee mode while I do some other work on a different miner

Can you run your numbers again with devfee turned on?

The miner will disable a number of optimizations if you disable devfee, resulting in slower hashrate.
member
Activity: 140
Merit: 18
I would strongly advise people to look at their mining pool hashrate for this miner and not the miners reported hashrate... Do the test...    Shocked

I'm beginning to think the same thing but it's more than that, when you first run it you get really nice looking numbers, then after a couple of hours, what you say sets in.

Here is the comparison of DSTM vs BMiner.

First line is DSTM that had been running for a few days
Next line is Bminer after 3 hours.  I saw grossly deviating current hash as well.

This was a GTX 950 that I just happened to have, but the two sets of number are on the same computer with the same hardware, etc.

Code:
Current Hashrate    Average Hashrate
      0H/s              155.9H/s
      80H/s             147.9H/s

As a final test, I'm going to leave it running in -nofee mode while I do some other work on a different miner
newbie
Activity: 39
Merit: 0
I would strongly advise people to look at their mining pool hashrate for this miner and not the miners reported hashrate... Do the test...    Shocked
hero member
Activity: 651
Merit: 501
My PGP Key: 92C7689C
Looks like after a few minutes, it's settling in around 1.68 kS/s on my rig:

* four 1070s: three MSI Gaming Xs and a PNY reference design
* Gentoo Linux
* mining at zpool

The MSI cards are power-limited to 115W and underclocked by 100 MHz.  The PNY card is power-limited to 95W and running at stock speed.  Total power consumption for the rig is 525W.

With the EWBF miner, I was seeing 1.52 kS/s.  tpruvot's ccminer was running closer to 1.2, so even with a devfee, it looks like you're still coming out ahead.
newbie
Activity: 19
Merit: 0
Just tried again for fun -- still doesn't start for me on Windows 10. Not from batch file, nor on its own with any command arguments.

And that's with two different Windows 10 machines, one AMD Ryzen 7 1700, one intel i7 3770. Both with a GTX 1070.

Any way to debug this? I don't have anything strange running at the same time, as far as I can tell.

Just to confirm -- did you turn off windows defender? Windows Defender sometimes would deny the execution.

What happens if you directly click on bminer.exe?

Yes, I had tried disabling Windows Defender, despite it showing no indication that it cared about bminer.exe. Disabling Windows Defender had no effect.

If I double-click bminer.exe a cmd-style window appears for a couple seconds, then closes. It has no contents. There are no error messages or other outputs.

EDIT: same result with new 5.2.0 version.

Try installing Visual Basic libraries, .NET Framework libraries, and latest CUDA SDK. 99% it will fix this for you

I do some very basic, minimal compiling on these machines, such as compiling xmrig & xmr-stak. So I have the CUDA SDK, and some .NET/VB libraries. Can you suggest specifics on which ones I might be missing?

What are the outputs if you are running the executable directly?

No output. When I run bminer.exe in cmd window, it just pauses, then goes back to the prompt. When I double-click it, an empty console window briefly appears, then disappears.
member
Activity: 461
Merit: 49
Hi,

This doesn't work on my rig. I'm using DSTM for two months and no issues.

Code:
rig2:~/bminer$ ./bminer -uri $SCHEME://$USERNAME@$POOL
[INFO] [2018-01-23T19:14:56-05:00] Bminer: When Crypto-mining Made Fast (v5.2.0-4f20af3)
[INFO] [2018-01-23T19:14:56-05:00] Starting miner on devices [0 1 2 3 4]
[INFO] [2018-01-23T19:14:56-05:00] Connected to us1-zcash.flypool.org:3333
[INFO] [2018-01-23T19:14:56-05:00] Subscribed to stratum server
[INFO] [2018-01-23T19:14:56-05:00] Set nonce to 108ddde5b3
[INFO] [2018-01-23T19:14:56-05:00] Authorized
[INFO] [2018-01-23T19:14:56-05:00] Set target to 6abc74931804560e2db29defa7c64b37894160e5d022dbf97e6abc7493180400
[INFO] [2018-01-23T19:14:56-05:00] Starting miner on device 0...
[INFO] [2018-01-23T19:14:56-05:00] [D0] GPU check watchdog has started
[INFO] [2018-01-23T19:14:57-05:00] Started miner on device 0
[INFO] [2018-01-23T19:14:57-05:00] Starting miner on device 1...
[INFO] [2018-01-23T19:14:57-05:00] [D1] GPU check watchdog has started
[INFO] [2018-01-23T19:14:57-05:00] Received new job 63d5ec62c08965e842d3
[INFO] [2018-01-23T19:14:57-05:00] Started miner on device 1
[INFO] [2018-01-23T19:14:58-05:00] Starting miner on device 2...
[INFO] [2018-01-23T19:14:58-05:00] [D2] GPU check watchdog has started
[INFO] [2018-01-23T19:14:59-05:00] Started miner on device 2
[INFO] [2018-01-23T19:14:59-05:00] Starting miner on device 3...
[INFO] [2018-01-23T19:14:59-05:00] [D3] GPU check watchdog has started
[INFO] [2018-01-23T19:15:00-05:00] Started miner on device 3
[INFO] [2018-01-23T19:15:01-05:00] Starting miner on device 4...
[INFO] [2018-01-23T19:15:01-05:00] [D4] GPU check watchdog has started
[INFO] [2018-01-23T19:15:02-05:00] Started miner on device 4
[WARN] [2018-01-23T19:15:02-05:00] Miner died! It will be restarted soon...

The dmesg shows me this message:

Code:
rig2:~# dmesg
[99684.245823] traps: bminer[22778] trap invalid opcode ip:6f6a92 sp:7fa0b8ffe930 error:0
[99944.004193] traps: bminer[22900] trap invalid opcode ip:6f6a92 sp:7fc424dfe930 error:0
[99944.004389] traps: bminer[22880] trap invalid opcode ip:6f6a92 sp:7fc4f9110930 error:0

How to resolve this?

Ubuntu 16.0.4
nvidia-387.26-0ubuntu1

What CPU are you using?
newbie
Activity: 12
Merit: 0
Hi,

This doesn't work on my rig. I'm using DSTM for two months and no issues.

Code:
rig2:~/bminer$ ./bminer -uri $SCHEME://$USERNAME@$POOL
[INFO] [2018-01-23T19:14:56-05:00] Bminer: When Crypto-mining Made Fast (v5.2.0-4f20af3)
[INFO] [2018-01-23T19:14:56-05:00] Starting miner on devices [0 1 2 3 4]
[INFO] [2018-01-23T19:14:56-05:00] Connected to us1-zcash.flypool.org:3333
[INFO] [2018-01-23T19:14:56-05:00] Subscribed to stratum server
[INFO] [2018-01-23T19:14:56-05:00] Set nonce to 108ddde5b3
[INFO] [2018-01-23T19:14:56-05:00] Authorized
[INFO] [2018-01-23T19:14:56-05:00] Set target to 6abc74931804560e2db29defa7c64b37894160e5d022dbf97e6abc7493180400
[INFO] [2018-01-23T19:14:56-05:00] Starting miner on device 0...
[INFO] [2018-01-23T19:14:56-05:00] [D0] GPU check watchdog has started
[INFO] [2018-01-23T19:14:57-05:00] Started miner on device 0
[INFO] [2018-01-23T19:14:57-05:00] Starting miner on device 1...
[INFO] [2018-01-23T19:14:57-05:00] [D1] GPU check watchdog has started
[INFO] [2018-01-23T19:14:57-05:00] Received new job 63d5ec62c08965e842d3
[INFO] [2018-01-23T19:14:57-05:00] Started miner on device 1
[INFO] [2018-01-23T19:14:58-05:00] Starting miner on device 2...
[INFO] [2018-01-23T19:14:58-05:00] [D2] GPU check watchdog has started
[INFO] [2018-01-23T19:14:59-05:00] Started miner on device 2
[INFO] [2018-01-23T19:14:59-05:00] Starting miner on device 3...
[INFO] [2018-01-23T19:14:59-05:00] [D3] GPU check watchdog has started
[INFO] [2018-01-23T19:15:00-05:00] Started miner on device 3
[INFO] [2018-01-23T19:15:01-05:00] Starting miner on device 4...
[INFO] [2018-01-23T19:15:01-05:00] [D4] GPU check watchdog has started
[INFO] [2018-01-23T19:15:02-05:00] Started miner on device 4
[WARN] [2018-01-23T19:15:02-05:00] Miner died! It will be restarted soon...

The dmesg shows me this message:

Code:
rig2:~# dmesg
[99684.245823] traps: bminer[22778] trap invalid opcode ip:6f6a92 sp:7fa0b8ffe930 error:0
[99944.004193] traps: bminer[22900] trap invalid opcode ip:6f6a92 sp:7fc424dfe930 error:0
[99944.004389] traps: bminer[22880] trap invalid opcode ip:6f6a92 sp:7fc4f9110930 error:0

How to resolve this?

Ubuntu 16.0.4
nvidia-387.26-0ubuntu1
member
Activity: 461
Merit: 49
bminer might report faster in the miner but on nanopool im seeing less than ewbf

first half of this is 1060 3gb on ewbf 2nd is with bminer


What commands did you use for tests?

What are the clocks and memory settings?

One common mistake is to disable devfee. As described in the documentation, disabling devfees will disable some of the optimizations which that results less hashrates on the pool side.
member
Activity: 461
Merit: 49
Just tried again for fun -- still doesn't start for me on Windows 10. Not from batch file, nor on its own with any command arguments.

And that's with two different Windows 10 machines, one AMD Ryzen 7 1700, one intel i7 3770. Both with a GTX 1070.

Any way to debug this? I don't have anything strange running at the same time, as far as I can tell.

Just to confirm -- did you turn off windows defender? Windows Defender sometimes would deny the execution.

What happens if you directly click on bminer.exe?

Yes, I had tried disabling Windows Defender, despite it showing no indication that it cared about bminer.exe. Disabling Windows Defender had no effect.

If I double-click bminer.exe a cmd-style window appears for a couple seconds, then closes. It has no contents. There are no error messages or other outputs.

EDIT: same result with new 5.2.0 version.

Try installing Visual Basic libraries, .NET Framework libraries, and latest CUDA SDK. 99% it will fix this for you

I do some very basic, minimal compiling on these machines, such as compiling xmrig & xmr-stak. So I have the CUDA SDK, and some .NET/VB libraries. Can you suggest specifics on which ones I might be missing?

What are the outputs if you are running the executable directly?
newbie
Activity: 8
Merit: 0
Goodminer so far.

For the people who are trying to add -logfile to version 5.2

Add this: -logfile log.txt  Wink
member
Activity: 297
Merit: 10
Note that I connected to Nicehash via miningrigrentals, not directly. That's where I get a lot of those rejected shares and errors. Setup a rig on miningrigrentals, add nicehash as pool to it, then connect bminer to miningrigrentals. The number of rejected shares shouldn't be different but bminer seems to get confused when nicehash changes nonces
newbie
Activity: 11
Merit: 0
Great work with the miner!
I've noticed the rejected shares go down drastically to about 1.2% on my machine anyways.
Pool side the hash rate looks sporadic, so its really hard to tell which miner averages a better over all rate.
I've got my 1070ti at about 550 sols/s and my 980 at about 330.
Compared to other miners I notice about 15-20 sols/s more between both cards, i'm enjoying it!
newbie
Activity: 106
Merit: 0
Anyone have this working on Titan XP cards? Works fine on 1080ti (about 4% faster than EWBF) but doesn't even start with Titans
newbie
Activity: 2
Merit: 0
Any advantages regarding Hashrate between 5.2 and 5.1? Actually i am running my 1070 Ti witch 500-510s, at60% PL, +260 Core, +595 Mem with 105-109 Power Consumption.
hero member
Activity: 562
Merit: 509
@cryptoyes
I'm getting less than 1% rejected shares.
member
Activity: 297
Merit: 10
5.2 seems to be worse ... I don't think 5.2 fixes the problem I reported earlier about Nicehash (and miningrigrentals). I still get lots of

Code:
Rejected share #152 ([1,"Share above target.",null])

and lots of

Code:
Rejected share #554 ([2,"Job not found.",null])

Sometimes 5-10 of them in a row. 5 rejected shares out of 85 is bad.

@realbminer: you may want to do some proper testing yourself on Nicehash, and on MiningRigRentals (with and without Nicehash as pool). Right now bminer is causing more harm than good when used on those.
newbie
Activity: 8
Merit: 0
Been using bMiner now for about a week after using EWBF first and then DSTM.

Very stable and getting the most sols out of the 3 miners from bMiner although DSTM is a close second.

My 1060 6gb is running at 335-345 (avg is 339 id say) sols with OC. And yea the reject rate is very low like someone mentioned earlier.

All in all, good job to the dev and thanks!
hav
newbie
Activity: 16
Merit: 0
it seems to be worse in 5.2
sometimes my sols just half.

Not sure if the devfee is secretly lots more or if this is a bug
Jump to: