Hiveon. №1 OS for mining

[gzf445]

i got a very strange problem.low hashrate on hiveos than win7
i used 4*gtx 750 to mine xmr and the hashrate change from 700h/s to 880h/s,the miner soft is xmr-stak.
while i use win7 the hashrate just from 860h-880h.the miner soft is xmr-stak,

Take a look at the Nvidia.txt config file that was used on Windows and compare it to what HiveOS generated. If HiveOS's Nvidia.txt config file is different you should put what Windows generated into the override field for HiveOS XMR-Stak.

Also, on Windows were the GTX 750's overclocked?

I have 24 GTX 750's installed as 8x GTX 750 on each of my three HP DL580 G7 servers. Even though all 24 GTX 750's are the exactly the same their hash rate varies from 221 to 250. The hash rate is steady for each GTX 750 it is just that one will produce 221 H/s and another will produce 250 H/s. The only thing I could guess at is that the memory speed on some is faster than others. On average these 24 GTX 750's produce 232 H/s each.

This is my Nvidia.txt for eight GTX 750:

"gpu_threads_conf" :
[
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 0,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 1,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 2,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 3,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 4,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 5,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 6,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 7,
    "threads" : 64, "blocks" : 4,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
],

hello.do you use the windows or hiveos.
cards have different hashrate because the different mem ram.
i have checked the nvidia.txt on windows and hiveos.
there is some different.
on windows like that

  { "index" : ,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
],

on hiveos like this

  { "index" : ,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
],

i have tried to change the nvidia.txt on hiveos but not worked.
it seems to reload everytime it runned.

[1714099058]

1714099058

[1714099058]

1714099058

[1714099058]

1714099058

[1714099058]

1714099058

[1714099058]

1714099058

[1714099058]

1714099058

[MinersRus]

hello.do you use the windows or hiveos.
cards have different hashrate because the different mem ram.
i have checked the nvidia.txt on windows and hiveos.
there is some different.
on windows like that

  { "index" : ,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 8, "bsleep" :  25,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
],

on hiveos like this

  { "index" : ,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : 0, "sync_mode" : 3,
  },
],

i have tried to change the nvidia.txt on hiveos but not worked.
it seems to reload everytime it runned.

You need to place the changes in your Wallet in the "Nvidia config override" field. HiveOS recreates the Nvidia.txt file from the Wallet which is why you are seeing it reloaded every time HiveOS starts.

I don't know how many GTX 750's you have but I have 8 and this is my Nvidia Config Override:

Code:

"gpu_threads_conf" :
[
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 0,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 1,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 2,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  { "index" : 3,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 4,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 5,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 2, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  { "index" : 6,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 6, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
  // gpu: GeForce GTX 750 architecture: 50
  //      memory: 938/1024 MiB
  //      smx: 4
  { "index" : 7,
    "threads" : 32, "blocks" : 12,
    "bfactor" : 7, "bsleep" :  0,
    "affine_to_cpu" : false, "sync_mode" : 3,
  },
],

You might also want to try "threads" : 64, "blocks" : 4 for each of the GPUs. Sometimes these values give higher hashes.

[HaloGenius]

Miners web update
- update fork of ccminer z-enemy v1.12 (use ccminer-update command to get latest versions of ccminer forks)
- NEW: t-rex v0.4c added as custom miner, find usage example at the bottom of the Wallets page

[N0pr3s3n7]

hey anyone knows why all of one card reports that it draws 75 W but the problematic card draws 96 W ?  I am using 8x gtx 1063 cards , fans are set fine , overclocking memory is set fine , but power readings always show that one of the cards draw a lot more power .

[Boriss]

hey anyone knows why all of one card reports that it draws 75 W but the problematic card draws 96 W ?  I am using 8x gtx 1063 cards , fans are set fine , overclocking memory is set fine , but power readings always show that one of the cards draw a lot more power .

You can try to start mining with only that card alone and see what happens, after that stop mining and try to start that card with one more.
Maybe settings are  not applied correctly on that specific card.

[bileq]

try reflash this more power consumed gpu. it should be fix this issue. you can download stable bios from one of other gpu and reflash from hive web interface.

[nitrobg]

hey anyone knows why all of one card reports that it draws 75 W but the problematic card draws 96 W ?  I am using 8x gtx 1063 cards , fans are set fine , overclocking memory is set fine , but power readings always show that one of the cards draw a lot more power .

If it's a Palit Dual, you'll need a newer bios. Follow this guide:

1) run this command: nvidia-smi --query-gpu=gpu_name,vbios_version,pci.device_id,pci.sub_device_id --format=csv
2) note the device ID (ABCD1234) and subdevice ID (EFGH5678)
3) reverse them by 4: 1234-ABCD-5678-EFGH
4) edit this link with the correct ID at the end: https://www.techpowerup.com/vgabios/?architecture=Uploads&did=1234-ABCD-5678-EFGH
5) download the newest bios and flash it

[N0pr3s3n7]

Thanks for advice guys I followed simply the guide on HIVE os forums to flash the vbios with nvflash_linux and everything worked as expected. I backed up my other same brand / type card vbios and flashed it to the other one and the problem is gone. I was also having a secondary trouble that card couldn't make it past 1000 mhz clock for etherum , but now it's working all fine. Thanks for the fast answers

[HaloGenius]

hey anyone knows why all of one card reports that it draws 75 W but the problematic card draws 96 W ?  I am using 8x gtx 1063 cards , fans are set fine , overclocking memory is set fine , but power readings always show that one of the cards draw a lot more power .

Palit with BIOS version  3a and 3b (at the end of BIOS version number) has this issue. Need newer BIOS as already suggest to you

GPU3 seems also Palit (7a - at the end of BIOS version number) and has newer BIOS without this issue

[dand070]

Has anyone been able to get EWBF 0.3 192_7 algo working for Zerocoin?  Seems I can connect but get rejected shares the entire time.

[fluxy12]

Hello, when will you install 11.9 claymore,

thank you

[dushan76]

Hey,
When do you plan to release V 2.0?

[HaloGenius]

Hello, when will you install 11.9 claymore,

thank you

from github commits 3 hours ago:
+ Add v11.9
- Remove v11.6

wait for new version

[HaloGenius]

Hey,
When do you plan to release V 2.0?

More than week ago we started private test.
Release of Hive 2.0 will hope by the end of the month.

[Warlockmng]

Hello guys, is someone mining Loki on HiveOS?

I cannot setup the miners for loki.hashvault.pro, can someone please help me?

Thanks in advance.

[HaloGenius]

HiveOS Linux GPU v0.5-63
- add Claymore Dual v11.9
- cpuminer-opt accepted/rejected report, small bugfix
- Autofan. Fixes for stability. New config option REBOOT_ON_ERROR=1, this will reboot rig if autofan can't read GPU temp or fan.
- Small preparations for Hive 2.0 release

[Boriss]

Thanks for advice guys I followed simply the guide on HIVE os forums to flash the vbios with nvflash_linux and everything worked as expected. I backed up my other same brand / type card vbios and flashed it to the other one and the problem is gone. I was also having a secondary trouble that card couldn't make it past 1000 mhz clock for etherum , but now it's working all fine. Thanks for the fast answers

Did you actually re-flashed all the cards to the latest BIOS or only that specific one that was problematic with power usage?
Would be interesting to see if that was their only fix or you could squeeze some more juice from the card with latest bios improvements

[grunoff]

HiveOS Linux GPU v0.5-63
- add Claymore Dual v11.9
- cpuminer-opt accepted/rejected report, small bugfix
- Autofan. Fixes for stability. New config option REBOOT_ON_ERROR=1, this will reboot rig if autofan can't read GPU temp or fan.
- Small preparations for Hive 2.0 release

Could you please add some more config options for autofan. Setting the max and min temps somehow would be great. Because at the moment I see that setting it by running the autofan.sh script, not really working.

[nightfury626]

i would like to use this OS but is the owner ever going to make xmrig mineable config for gpu's. i dont get the point of making it only cpu when its better than xmrstak

[HaloGenius]

i would like to use this OS but is the owner ever going to make xmrig mineable config for gpu's. i dont get the point of making it only cpu when its better than xmrstak

It's very simple - disable GPU mining by
put to AMD config
"gpu_threads_conf" : [], "platform_index" : 0,

put to NVidia config
"gpu_threads_conf" : [], "platform_index" : 1,

and configure threads for CPU mining (in following example 4 threads)
"cpu_threads_conf" : [
{ "low_power_mode" : false, "no_prefetch" : false, "affine_to_cpu" : false },
{ "low_power_mode" : false, "no_prefetch" : false, "affine_to_cpu" : false },
{ "low_power_mode" : false, "no_prefetch" : false, "affine_to_cpu" : false },
{ "low_power_mode" : false, "no_prefetch" : false, "affine_to_cpu" : false },
],

Also my suggestion use for cpu mining only - XMRig