miniZ v2.0a Equihash 144,5 125,4 210,9 150,5 192,7 BeamHash3 ProgPoW Ethash CFX - page 36.

miniZ

member

Activity: 690

Merit: 17

Quote from: $@X on September 18, 2019, 08:05:13 PM

Is anyone else seeing high CPU usage? I have 70% + being used on a quad core, for example. Mining BeamAlgoII

Can't post an image as I apparently still do not have enough cred Cool

On the bright side, the latest version has been completely stable for me and I see maybe 1 rejected share every few days. Great work!

Hi $@X,
thank you for letting us know.
Are you experiencing higher CPU usage in the latest version, when compared to the previous one? Is it happening only with Beam algorithm?
We expected this to be solved already.
Did you noticed a difference with latest driver, as suggested by joseph32?
We keep looking into this.
Cheers

crypt_fr

newbie

Activity: 21

Merit: 0

Quote from: miniZ on September 13, 2019, 09:41:50 PM

Hi everyone,
Thank you for your feedback!

You can find miniZ v1.5q6 here.
We made improvements to address some of the issues you have reported.

In brief:
* Performance improvement for 192,7. (Up to 4% on RTX)
* Added --jobtimeout, --retries, --retrydelay for (re)connection customization.
* Added --cleanjobs for miner to only accept clean jobs.
* Better connection handling.
* Bug fixes.

For additional information check our Usage or FAQ pages.
Let us know how it goes!

Happy mining!

Great job on the perf improvement on 192,7 for both RTX and GTX. I run some RTX 2060 Super and a couple GTX 1660 Ti and the + 4% is real.

This is some serious work you've done here. Kudos!!

PS: didn't see the info about this release on your Twitter

cnminer

newbie

Activity: 16

Merit: 0

no http or socket api interface , we can't get miner speed, shares, rejected... etc.

joseph32

member

Activity: 413

Merit: 21

If you see high CPU usage, go back to 430.86. Seems a bug in the newest driver.

I tried to update some 7 and 13 GPU rigs and all got instant 100% CPU load with that crap driver. Back to 430.86 and everything is fine again.

fr4nkthetank

legendary

Activity: 2294

Merit: 1182

Now the money is free, and so the people will be

I guess nicehash fixed their shit, I havent seen another worker error in a while.
I see 40-50% cpu usage on G4400 dual core for 6 1080tis. 70% seems high on a quad core if on all cores

$@X

newbie

Activity: 19

Merit: 0

Is anyone else seeing high CPU usage? I have 70% + being used on a quad core, for example. Mining BeamAlgoII

Can't post an image as I apparently still do not have enough cred Cool

On the bright side, the latest version has been completely stable for me and I see maybe 1 rejected share every few days. Great work!

fr4nkthetank

legendary

Activity: 2294

Merit: 1182

Now the money is free, and so the people will be

yeah thats pretty much my .bat file setup. I've increased the timeout value a bit (it was 1), see if that helps ill let you know

miniZ

member

Activity: 690

Merit: 17

Quote from: joseph32 on September 16, 2019, 11:17:06 AM

Quote from: fr4nkthetank on September 16, 2019, 08:45:29 AM

I put a gpu on nicehash since they still owe me a little coin from the hack a while ago, and got "worker not authorized". Miniz tries to restart the gpus and connection but gets stuck. If I manually close miniz and restart it, it connects. A workaround could be to have miniz shut down when it receives that error so the goto loop in the .bat file could restart the miniz ? If you need I could post a screenshot, i remember getting this error back in the early days of Zcash

Yes, its very annoying these days. Here is the error:

https://imgur.com/a/j8LcBWL

Problem is, in AwesomeMiner I cant catch that error fast enough because Miniz still prints the last Hashrate over the API. So it looks like hashing, but without new accepted shares.

Hi joseph32,
thanks for reporting this.
We will investigate this in AwesomeMiner.
For the next version we will also make sure the miner reports zero hashrate, or a different value, for AwesomeMiner to understand what happened.
Cheers

miniZ

member

Activity: 690

Merit: 17

Quote from: fr4nkthetank on September 16, 2019, 08:45:29 AM

I put a gpu on nicehash since they still owe me a little coin from the hack a while ago, and got "worker not authorized". Miniz tries to restart the gpus and connection but gets stuck. If I manually close miniz and restart it, it connects. A workaround could be to have miniz shut down when it receives that error so the goto loop in the .bat file could restart the miniz ? If you need I could post a screenshot, i remember getting this error back in the early days of Zcash

Hi fr4nkthetank,
thanks for the feedback.

miniZ should exit when that message happens. We will have a look.

Is your .bat file similar to this? :

Code:

:loop
miniZ.exe --url=YourWallet.worker@server:port --log --extra
timeout /t 5
goto loop

Cheers

miniZ

member

Activity: 690

Merit: 17

Quote from: Divinity666 on September 14, 2019, 11:21:33 AM

Well, I guess --jobtimeout=600(2x default) has finally solved my problems on 2miners pool. I'm mining with v1.5q6 for almost 24hours now and cards never went into idle so far.

Hi Divinity666,
thanks for letting us know!
Cheers

joseph32

member

Activity: 413

Merit: 21

Quote from: fr4nkthetank on September 16, 2019, 08:45:29 AM

I put a gpu on nicehash since they still owe me a little coin from the hack a while ago, and got "worker not authorized". Miniz tries to restart the gpus and connection but gets stuck. If I manually close miniz and restart it, it connects. A workaround could be to have miniz shut down when it receives that error so the goto loop in the .bat file could restart the miniz ? If you need I could post a screenshot, i remember getting this error back in the early days of Zcash

Yes, its very annoying these days. Here is the error:

https://imgur.com/a/j8LcBWL

Problem is, in AwesomeMiner I cant catch that error fast enough because Miniz still prints the last Hashrate over the API. So it looks like hashing, but without new accepted shares.

fr4nkthetank

legendary

Activity: 2294

Merit: 1182

Now the money is free, and so the people will be

I put a gpu on nicehash since they still owe me a little coin from the hack a while ago, and got "worker not authorized". Miniz tries to restart the gpus and connection but gets stuck. If I manually close miniz and restart it, it connects. A workaround could be to have miniz shut down when it receives that error so the goto loop in the .bat file could restart the miniz ? If you need I could post a screenshot, i remember getting this error back in the early days of Zcash

Divinity666

jr. member

Activity: 312

Merit: 2

Well, I guess --jobtimeout=600(2x default) has finally solved my problems on 2miners pool. I'm mining with v1.5q6 for almost 24hours now and cards never went into idle so far.

fr4nkthetank

legendary

Activity: 2294

Merit: 1182

Now the money is free, and so the people will be

Thanks for the update, looking forward to more stable connections

Although as people have said before a :start // goto start in the .bat file works wonders

miniZ

member

Activity: 690

Merit: 17

Hi everyone,
Thank you for your feedback!

You can find miniZ v1.5q6 here.
We made improvements to address some of the issues you have reported.

In brief:
* Performance improvement for 192,7. (Up to 4% on RTX)
* Added --jobtimeout, --retries, --retrydelay for (re)connection customization.
* Added --cleanjobs for miner to only accept clean jobs.
* Better connection handling.
* Bug fixes.

For additional information check our Usage or FAQ pages.
Let us know how it goes!

Happy mining!

ARTRN

newbie

Activity: 63

Merit: 0

Quote from: miniZ on September 06, 2019, 10:17:00 AM

Quote from: ARTRN on September 04, 2019, 08:56:28 AM

Just an FYI, v1.5q5 still lost connection 'can't connect to any pool'. I lost 45 mins mining time.

Hi ARTRN,
we have been checking all connection errors.
It seems that you lost connection to all your pools.
Could you let us know which coin are you mining and to which pool?
This can help us to further check if there can be any issue.
Thanks!
Cheers

Sorry I wasn't clear enough. It was on BEAM @sunpool, v1.5q5 cuda10 windows 10.
Thx.

miniZ

member

Activity: 690

Merit: 17

Quote from: V_I_T_A_L_Y on September 06, 2019, 04:58:12 AM

Hello! please tell me I'm your miner can dig beam on the pool solo-beam.2miners.com? And then yesterday switched to your miner Mini-Z with gminer and still have not found a single block, maybe just bad luck but decided better to ask you. Not all miners are suitable for solo mining

Hi V_I_T_A_L_Y,
miniZ should be able to mine solo on 2miners. We know it can find blocks on other pools.
Also, 2miners accepts all our shares when mining solo.

Quote from: V_I_T_A_L_Y on September 06, 2019, 06:14:52 AM

Found the unit and the rejected, coincidence?

Do you mean you get rejected shares? In this case could you let us know which miniZ version are you using?
Or do you mean that a block was found and rejected. Unfortunately this can happen in case of orphan block.
Maybe this can help to understand it better?

Thanks.
Cheers

miniZ

member

Activity: 690

Merit: 17

Quote from: BigHarryArmsi on September 04, 2019, 10:02:24 AM

More speed!

Thanks again, miniZ

update... I must say the 1066s are very responsive to power increase. Increasing power from 65% to 75% gained 10% in hash rate!
up-update... invalids have dropped to near non-existent!

Hi BigHarryArmsi,
good to know the invalid shares issue is fixed.
Yes, raising power a little has large effect on performance.
It is good to adjust to have better efficiency, min power but high hashrate.

We always work for speed improvements, but we are also focused on issues people have been reporting.

Thanks for the feedback. Keep us posted with your results.
Cheers

miniZ

member

Activity: 690

Merit: 17

Quote from: darkneorus on September 04, 2019, 09:09:32 AM

found this at your website:

Quote

Some of you were unable to run miniZ with the NVIDIA mining series cards, on Windows. These cards require an older driver (≥382.53, depending on card and brand) that is not compatible with cuda 9.2 (see Table 1).

my P106 and P104 are running with 4xx drivers and CUDA 10 without any issues. maybe you have some hacks/workarounds in your code to support CUDA 8 for them, and they are related to, or are the cause of low performance?

thank you

Hi darkneorus,
that information is outdated. We have removed it.
Thank you for the suggestion, but it does not seem to be related to it.
As we said before, since we do not have P104/6 it is hard to optimise the performance for those GPUs.
We'll keep doing our best.
Cheers

miniZ

member

Activity: 690

Merit: 17

Quote from: ARTRN on September 04, 2019, 08:56:28 AM

Just an FYI, v1.5q5 still lost connection 'can't connect to any pool'. I lost 45 mins mining time.

Hi ARTRN,
we have been checking all connection errors.
It seems that you lost connection to all your pools.
Could you let us know which coin are you mining and to which pool?
This can help us to further check if there can be any issue.
Thanks!
Cheers

Topic: miniZ v2.0a Equihash 144,5 125,4 210,9 150,5 192,7 BeamHash3 ProgPoW Ethash CFX - page 36. (Read 59613 times)