[ANN] dstm's ZCash / Equihash Nvidia Miner v0.6.2 (Linux / Windows)

[akuci]

Woooohoooo! Hell Yeah DSTM! Will test it out right now.

[1714543774]

1714543774

[1714543774]

1714543774

[1714543774]

1714543774

[1714543774]

1714543774

[1714543774]

1714543774

[1714543774]

1714543774

[rwaters]

Thank you for version 0.6.0!

Still no stratum "client.reconnect" support? You mentioned it is in the works in November :/
From the Stratum mining protocol info: client.reconnect("hostname", port, waittime). The client should disconnect, wait waittime seconds (if provided), then connect to the given host/port (which defaults to the current server).


When trying to connect to miningrigrentals.com, it just loops through this failure:

Code:

unknown method client.reconnect
#  recv failed: 10054

ewbf seems to work, would really like to switch to dstm though.

[skablast]

DSTM, I know you are the  kind of guy, do it once, do it well.
But time has passed, and the difference in performance on the poor man rigs with 1050ti with other miner, (excavator) is increasing.
As the ROI on 1080 and 1070 is decreasing with the actual crazy price, we are looking more closely to lower spec that offer quicker ROI
I remember you adapted some 1060 sw to it, please give come back and give it a look, as I'm an old guy that doesn't like to change  what is working well.
thanks from someone that's just paying for your morning coffee.

[thecow]

Anyone else getting "Color mode not supported" in Windows 7 64-Bit? My Win10 64 has color.

Otherwise the hashrate is the same, I guess 800Sol/s is wishful thinking.

[sir_blacks]

Anyone else getting "Color mode not supported" in Windows 7 64-Bit? My Win10 64 has color.

Otherwise the hashrate is the same, I guess 800Sol/s is wishful thinking.

- i can confirm in Windows 7 Service Pack 1 with "--color 1" = noo color,
- speed speaking my 1060 it get slice better with tune what i use PL 74% Core Clock +168 Mem Clock +600 aka from Avg ~301 it goo to ~306 but this it require more time to be sure
- 1 crash with 0.6 till now after using 0.5.8 with 0 crash

[Biggen1]

I'll let you guys beta test. 0.5.8 works too good for me on Linux to screw around with.

[Hotmetal]

I'll let you guys beta test. 0.5.8 works too good for me on Linux to screw around with.

I'm testing on a bunch of identical Nvidia 1070 rigs, will post back feedback.

[Hotmetal]

I'm testing on a bunch of identical Nvidia 1070 rigs, will post back feedback.

v0.6 is pretty solid on my 1070's so I'll be rolling it out to my farm shortly.

[jpl]

New Version 0.6

changelog
- failover pool support
- introduce configuration file
- introduce intensity option
- support per gpu temp-target via cfg-file
- support per gpu intensity via cfg-file
- json-rpc report gpu_name, pci_bus_id, pci_device_id
  gpu_uuid
- improve error reporting
- improve handling of invalid server resposes
- improve reconnection in cases where servers
  become reachable but respond with malformed data
- ui: add colorized output
- web-ui: use dark colors


0.6 contains a huge amount of changes, there are especially quite a lot internal changes - so please test before deploying.
Most notable in this release is failover pool support and the addition of configuration files. You can setup multiple pools in the configuration file. Configured pools are ordered by priority. ZM connects to the highest priority pool which is reachable. If a pool gets unreachable a lower priority pool will be used. ZM reconnects to a higher priority pool automatically if it gets reachable again, accepts your login and responds properly. A bad behaving server (which oscillates between being responsive / unresponsive states) might cause zm to oscillate between the servers - I'll look for ways to improve this in future releases.

There is also a way to reduce the load which is put on the gpu now. It's configured by the intensity option. It takes some time (about 30sec) till it starts reducing the gpu load since it needs to gather performance data first. The intensity option accepts values in the range ]0.0 - 1.0[. A value of 0.5 will roughly put a load of 50%.  

0.6 has also some ui changes. It supports colorized terminal output, webui uses dark colors now, since this was requested multiple times.

There are ofc multiple internal improvements all around the place, most notable better handling of invalid data and improved error reporting.

Nice!  Thanks for all your hard work dstm.  (I've developed apps for iPhone, ..and its a pain in the ass, so I'm sure this is a bit of a task also)

[Andrey09]

DSTM, could you make as option --white or --black web Ui? I guess white was better, All colors which we need in miner windows it's only GPU temps and total hashrate.

Gpu temps colors:

0-70 C - green color
70-80 C - ellow color
80-100 C - red color

We don't need all text as colored right now, just GPU temps and total information about speed, and may be Uptime in console window.

Tnx you.

[crypto4pizza]

Just switched over from EWBF.  Is there a temp limit option?  I only see an option for "temp-target", but this doesn't look like the same thing unless i'm mistaken.

[cozyfire]

Just switched from EBWF too, however I'm having multiple stability issues, the miner tells me to lower the overclock.

Can anyone share their settings with me?, right now I have two rigs, both with 6x gtx 1070, had to lower the oc to +100/+500 due to this (previously +150/+600), however I think I can increase memory more.

I use +100 +300 + autooverclock in EVGA precision on my EVGA FTW 1070ti it works great ~500Sol

[cozyfire]

Just switched from EBWF too, however I'm having multiple stability issues, the miner tells me to lower the overclock.

Can anyone share their settings with me?, right now I have two rigs, both with 6x gtx 1070, had to lower the oc to +100/+500 due to this (previously +150/+600), however I think I can increase memory more.

I use +100 +300 + autooverclock in EVGA precision on my EVGA FTW 1070ti it works great ~500Sol

I think something might be wrong with one of my cards, it barely reach 430 sols/s

Disable all overclocking. Start DSTM and mine for like 8-12 hours, enable logs. Then add some OC and repeat until satisfied. In my experience DSTM doesn't like overclocking much, proceed with care, also look closely at shares per minute - its more important than hashrate IMO

[malthrax]

New Version 0.6

me likey!

[neekir]

This is the result , I am getting in 1070 and 1060 6GB

GPU0  66C  Sol/s: 453.4  Sol/W: 3.80  Avg: 458.3  I/s: 243.3  Sh: 1.99   1.00 297 +
>  GPU1  66C  Sol/s: 336.7  Sol/W: 3.07  Avg: 339.8  I/s: 180.5  Sh: 0.99   1.00 268
>  GPU2  62C  Sol/s: 347.1  Sol/W: 3.12  Avg: 343.6  I/s: 182.7  Sh: 4.98   1.00 266 +

1060 3GB (hynix)

   GPU0  77C  Sol/s: 262.5  Sol/W: 3.39  Avg: 278.6  I/s: 139.6  Sh: 1.79   1.00 250
   GPU1  77C  Sol/s: 277.9  Sol/W: 3.23  Avg: 289.5  I/s: 150.5  Sh: 0.30   1.00 238
   GPU2  77C  Sol/s: 273.7  Sol/W: 3.29  Avg: 287.9  I/s: 145.5  Sh: 1.19   1.00 235
   GPU3  77C  Sol/s: 252.5  Sol/W: 3.46  Avg: 272.2  I/s: 133.4  Sh: 0.30   1.00 234
   GPU4  77C  Sol/s: 262.1  Sol/W: 3.36  Avg: 278.7  I/s: 142.5  Sh: 1.19   1.00 235
   GPU5  77C  Sol/s: 279.1  Sol/W: 3.28  Avg: 289.4  I/s: 150.0  Sh: 1.49   1.00 247
   GPU6  77C  Sol/s: 276.3  Sol/W: 3.19  Avg: 287.5  I/s: 149.1  Sh: 0.30   1.00 247
   ========== Sol/s: 1884.1 Sol/W: 3.31  Avg: 1983.9 I/s: 1010.4 Sh: 6.57   1.00 240

[dvince]

2% fee is deducted from total hashrate or is it already included in total hashrate calculation?

[akuci]

This is the result , I am getting in 1070 and 1060 6GB

GPU0  66C  Sol/s: 453.4  Sol/W: 3.80  Avg: 458.3  I/s: 243.3  Sh: 1.99   1.00 297 +
>  GPU1  66C  Sol/s: 336.7  Sol/W: 3.07  Avg: 339.8  I/s: 180.5  Sh: 0.99   1.00 268
>  GPU2  62C  Sol/s: 347.1  Sol/W: 3.12  Avg: 343.6  I/s: 182.7  Sh: 4.98   1.00 266 +

1060 3GB (hynix)

   GPU0  77C  Sol/s: 262.5  Sol/W: 3.39  Avg: 278.6  I/s: 139.6  Sh: 1.79   1.00 250
   GPU1  77C  Sol/s: 277.9  Sol/W: 3.23  Avg: 289.5  I/s: 150.5  Sh: 0.30   1.00 238
   GPU2  77C  Sol/s: 273.7  Sol/W: 3.29  Avg: 287.9  I/s: 145.5  Sh: 1.19   1.00 235
   GPU3  77C  Sol/s: 252.5  Sol/W: 3.46  Avg: 272.2  I/s: 133.4  Sh: 0.30   1.00 234
   GPU4  77C  Sol/s: 262.1  Sol/W: 3.36  Avg: 278.7  I/s: 142.5  Sh: 1.19   1.00 235
   GPU5  77C  Sol/s: 279.1  Sol/W: 3.28  Avg: 289.4  I/s: 150.0  Sh: 1.49   1.00 247
   GPU6  77C  Sol/s: 276.3  Sol/W: 3.19  Avg: 287.5  I/s: 149.1  Sh: 0.30   1.00 247
   ========== Sol/s: 1884.1 Sol/W: 3.31  Avg: 1983.9 I/s: 1010.4 Sh: 6.57   1.00 240

Your 3GB 1060s are WAY TO HOT for 24\7. Where are you keeping them?
Also your results are too low for a 3GB 1060, something is very wrong with your setup.

[promomei]

This is the result , I am getting in 1070 and 1060 6GB

GPU0  66C  Sol/s: 453.4  Sol/W: 3.80  Avg: 458.3  I/s: 243.3  Sh: 1.99   1.00 297 +
>  GPU1  66C  Sol/s: 336.7  Sol/W: 3.07  Avg: 339.8  I/s: 180.5  Sh: 0.99   1.00 268
>  GPU2  62C  Sol/s: 347.1  Sol/W: 3.12  Avg: 343.6  I/s: 182.7  Sh: 4.98   1.00 266 +

1060 3GB (hynix)

   GPU0  77C  Sol/s: 262.5  Sol/W: 3.39  Avg: 278.6  I/s: 139.6  Sh: 1.79   1.00 250
   GPU1  77C  Sol/s: 277.9  Sol/W: 3.23  Avg: 289.5  I/s: 150.5  Sh: 0.30   1.00 238
   GPU2  77C  Sol/s: 273.7  Sol/W: 3.29  Avg: 287.9  I/s: 145.5  Sh: 1.19   1.00 235
   GPU3  77C  Sol/s: 252.5  Sol/W: 3.46  Avg: 272.2  I/s: 133.4  Sh: 0.30   1.00 234
   GPU4  77C  Sol/s: 262.1  Sol/W: 3.36  Avg: 278.7  I/s: 142.5  Sh: 1.19   1.00 235
   GPU5  77C  Sol/s: 279.1  Sol/W: 3.28  Avg: 289.4  I/s: 150.0  Sh: 1.49   1.00 247
   GPU6  77C  Sol/s: 276.3  Sol/W: 3.19  Avg: 287.5  I/s: 149.1  Sh: 0.30   1.00 247
   ========== Sol/s: 1884.1 Sol/W: 3.31  Avg: 1983.9 I/s: 1010.4 Sh: 6.57   1.00 240

Your 3GB 1060s are WAY TO HOT for 24\7. Where are you keeping them?
Also your results are too low for a 3GB 1060, something is very wrong with your setup.

blame hynix memory

[od1n]

Anyone else getting "Color mode not supported" in Windows 7 64-Bit? My Win10 64 has color.

Same here with Win7: "Color mode not supported"  

[sir_blacks]

@neekir your 1060 are not working good on 250-277 Sol

a. mining on 250 mls latency = BAD not know on what pool you are but better try to get a better one with 30-60 mls at list
b. i am tenting to believe your I/s (intensity), low one, it cause by High temperature aka miner it lower to keep temp lower ... with it can cause low hrate aka that 250-280 Sol/s
c. 77 C temp is to high try to fix cooling somehow ... it will improve point b

- equihash algoritm it not care much about memory speed it need more brute force from GPU combine by cuda core & pipeline soo hynix i think is not a realy a issue
- try to post technical data use in tuning like Power Limit, Temp Limit, GPU clock, Mem Clock

@akuci yup for sure he have some bad thing in his rig ... @promomei equihash algoritm it not care much about memory speed

for 1060 with PL 73%, TL 80, GPU Clock +168 (1830 - 1880 Mhz), Mem Clock +600 (4404 mhz) it suppose to be something like this (mining btg):

Code:

2018-02-16 15:24:49|   GPU0  61C  Sol/s: 301.1  Sol/W: 3.46  Avg: 302.1  I/s: 162.1  Sh: 2.22   1.00 47  ++
2018-02-16 15:25:09|   GPU0  61C  Sol/s: 303.3  Sol/W: 3.46  Avg: 302.1  I/s: 161.8  Sh: 2.22   1.00 47  +
2018-02-16 15:25:30|>  GPU0  61C  Sol/s: 309.0  Sol/W: 3.46  Avg: 302.1  I/s: 161.8  Sh: 2.21   1.00 47  
2018-02-16 15:25:50|   GPU0  61C  Sol/s: 302.3  Sol/W: 3.46  Avg: 302.1  I/s: 161.9  Sh: 2.22   1.00 47  ++
2018-02-16 15:26:10|   GPU0  61C  Sol/s: 302.5  Sol/W: 3.46  Avg: 302.1  I/s: 161.9  Sh: 2.22   1.00 47