Pages:
Author

Topic: excavator by NiceHash - multi-algorithm NVIDIA GPU miner [1.5.6a] - page 10. (Read 179296 times)

jr. member
Activity: 132
Merit: 7
Which is better this or hsr?

I'm assuming you're asking about the Neoscrypt. Well.. It depends on the card Wink
jr. member
Activity: 132
Merit: 7
lol new version 2.0.1.3 removed daggerhashimoto for 3gb cards!!! LOL they dont know till April is profitable?

We decided to drop 3GB cards for daggerhashimoto because the DAG size was getting dangerously close to the 3GB limit and daggerhashimoto didn't even run on these cards especially when they are used as primary cards. This may not be a problem for experienced miners, but the majority of NHM users are casual ones. There are so many things to consider, you know...

I have been trying to figure out why I had no daggerhashimoto on my GTX 1060 3gb rig with NHM. now I know.
But this site below says the dag file is only 2.31gb yet, nowhere near the 3 gb limit ?

https://investoon.com/tools/dag_size

As mentioned before we dropped the daggerhashimoto for 3gb cards because there were a lot of issues/crashes with it among NHM users (especially on primary cards). We could remove the 3gb condition in excavator though, and you can run it without NHM then..
jr. member
Activity: 132
Merit: 7
As promised I uploaded a simple benchmark script for excavator - https://github.com/nicehash/excavator/blob/master/excavator-benchmark.py

Python 3 is required!

Examples:

1. benchmark algorithm 1 on device 0
python excavator-benchmark.py --excavator path_to_excavator -a algo1 -d 0 -b

2. benchmark all algorithms on all devices
python excavator-benchmark.py --excavator path_to_excavator -b

3. benchmark on nicehash pool (default is eu server)
python excavator-benchmark.py --excavator path_to_excavator -u btc_address

Default time for benchmark is 300 seconds per algorithm, you can modify it with -t parameter.



Any way to edit the script to use multiple workers?  It would be nice if this could be dynamic based on the algo.

Workers are added at line 210. You can copy the append command multiple times or add a for loop to run multiple workers per device.

Example #1: To add two workers per device change

Code:
for dev in devices:
      workers.append(self.add_worker(alg_id, dev[0], algo[1:]))

to

Code:
for dev in devices:
      workers.append(self.add_worker(alg_id, dev[0], algo[1:]))
      workers.append(self.add_worker(alg_id, dev[0], algo[1:]))

To make it algorithm specific you could always check which algorithm is processed at the moment.

Example #2: Add two workers only when running equihash:
Code:
for dev in devices:
      workers.append(self.add_worker(alg_id, dev[0], algo[1:]))
      if algo[0] == "equihash":
            workers.append(self.add_worker(alg_id, dev[0], algo[1:]))
hero member
Activity: 687
Merit: 511
Any update on when you'll be making the next Linux release (ideally with the fan speed fix)?  You're killing me with all of these Windows releases without Linux builds... Wink
newbie
Activity: 106
Merit: 0
Regarding those neoscrypt speed improvements...I'm seeing almost 2x. Is that real? says Titan XP is now doing 2.14MH?!
newbie
Activity: 28
Merit: 0
Just started testing 1.4.3a with my 5 x nvidia 1080ti rig. It seems way more stable than 1.1.0a. I knew excavator had potential against EWBF, this is why I'm giving another try.

What I love is that the GPUs are rather balanced in terms of output (680h/s) and total hashrate doesn't fluctuate much as in EWBF: Last 300 shares are all submitted  around 3.38kh/s.
 
What I don't like though, is that the sum is below 3500 which is the average at EWBF (3400min / 3600 max)

It stopped printing the results after 20 minutes but my readings from the pool look ok.

I wonder how I can set thread and block sizes via the .json file? I'm not sure if excavator is using all the available juice :/

EDIT: I'm mining Zcash "equihash"
sr. member
Activity: 518
Merit: 250
I refuse to use Excavator for anything other than Nicehash. I get better results from the latest cudaminer. Plus, I have a theory that all those "benchmarks" that nicehash requires when a mosquito lands on your GPU are giving them free hashes.

The reason i stopped use the nicehash miner 2 is becourse of all those god damn benchmarks. Every time i restart the god damn program i have 2 benchmark eight fucking GPUs?! It takes forever, like 5 hours. Thats alot in mining time.
newbie
Activity: 32
Merit: 0
I refuse to use Excavator for anything other than Nicehash. I get better results from the latest cudaminer. Plus, I have a theory that all those "benchmarks" that nicehash requires when a mosquito lands on your GPU are giving them free hashes.
newbie
Activity: 77
Merit: 0
Which is better this or hsr?
member
Activity: 118
Merit: 10
lol new version 2.0.1.3 removed daggerhashimoto for 3gb cards!!! LOL they dont know till April is profitable?

We decided to drop 3GB cards for daggerhashimoto because the DAG size was getting dangerously close to the 3GB limit and daggerhashimoto didn't even run on these cards especially when they are used as primary cards. This may not be a problem for experienced miners, but the majority of NHM users are casual ones. There are so many things to consider, you know...

I have been trying to figure out why I had no daggerhashimoto on my GTX 1060 3gb rig with NHM. now I know.
But this site below says the dag file is only 2.31gb yet, nowhere near the 3 gb limit ?

https://investoon.com/tools/dag_size
newbie
Activity: 11
Merit: 0
As promised I uploaded a simple benchmark script for excavator - https://github.com/nicehash/excavator/blob/master/excavator-benchmark.py

Python 3 is required!

Examples:

1. benchmark algorithm 1 on device 0
python excavator-benchmark.py --excavator path_to_excavator -a algo1 -d 0 -b

2. benchmark all algorithms on all devices
python excavator-benchmark.py --excavator path_to_excavator -b

3. benchmark on nicehash pool (default is eu server)
python excavator-benchmark.py --excavator path_to_excavator -u btc_address

Default time for benchmark is 300 seconds per algorithm, you can modify it with -t parameter.



Any way to edit the script to use multiple workers?  It would be nice if this could be dynamic based on the algo.
jr. member
Activity: 40
Merit: 1
Is there a way to use excavator without .json files? Just the batch file like other miners like ccminer, sgminer or claymore
newbie
Activity: 49
Merit: 0
for who are confuse about TPD & Real W Power & GPU Clock Ration i made a picture

https://i.imgur.com/ejH72Yp.png
https://i.imgur.com/ejH72Yp.png

cheers

Oh Sad You complicate everything

If you need "Power Limit" MSI Afterburner 72% in Excavator you must write:
{"id":1,"method":"device.set.tdp","params":["0","72"]}

MSI Afterburner "Power Limit" = Excavator "device.set.tdp" understandably ?


Agree !

I would add that, as an old school overclocker, I don't use TDP/Power/Temp limits as adjustment variables but only as a security.
If you set the proper Core Freq / Voltage to reach stability you won't be bothered with sampling issues of the power limiter.
jr. member
Activity: 119
Merit: 3
for who are confuse about TPD & Real W Power & GPU Clock Ration i made a picture

https://i.imgur.com/ejH72Yp.png


cheers
jr. member
Activity: 119
Merit: 3
@DumaxFr api it have 2 option power_limit = max power in W or tdp in % booth are yes relative value no GPU it will stay on exact value what it will be set & yes it will be +/- variation, but problem is when an application can't fallow his own setting & exceed to much like in my case, for that i will give you few scenario below:

- GPU TDP = 72% = ~86W = ~1860 Mhz = good ration = normal working
- GPU TPD = 72% = ~86W = ~2216 Mhz = miner it will crash & GPU it will to
- GPU TDP = 72% = ~88W = ~1888 MHz fine in range of max with +/-
- GPU TDP = 72% = ~90W = ~1847 Mhz = Bad aka application exceed limit result escape out of control

anyway no matter what Excavator it do something what make GPU to hit max with bad power ratio & ending booth screw up, so back to my point Excavator it need to be tech to be "good boy" because 1. it can fry equipment 2. it end not doing his job 3. it consume power for nothing

@Lexabush is a difference between that & yes @DumaxFr i read it before i use it Wink
A. device.set.power_limit = string = New power limit in Watts.

https://github.com/nicehash/excavator/tree/master/api#-devicesetpower_limit
- so you set real W power what you want to be max
Code:
{"id":1,"method":"device.set.power_limit","params":["0","86"]}

B. device.set.tdp = string = New TDP limit %.
https://github.com/nicehash/excavator/tree/master/api#-devicesettdp

- so you set % of Power what you want to be use
Code:
{"id":1,"method":"device.set.tdp","params":["0","72"]}
newbie
Activity: 49
Merit: 0
because usual i like to find a fix to a problem, after reading more about https://github.com/nicehash/excavator/tree/master/api i was say hey let try to use "device.set.power_limit" setting my be i will make Excavator to be stable & work nice agen!!! so i set 86W limit (this is my target set already in Afterburner) to be sure it will never exceed that value & GPU Clock it will never goo more that 1860 Mhz. SURPRISE Excavator FAIL to obey / fallow his OWN fraking setting ending crashing GPU clock (2088 - 2216) => ending crashing himself !!! Way ?!?!? i will like to undestend but that's dev job to look in code & see way is out of control, way he can't fallow his own setting ... way can't be stable

Setting extra add in json:
Code:
{"time":5,"commands":[
{"id":1,"method":"device.set.power_limit","params":["0","86"]}
]},

i make a image to show working stable miner & unstable one on crash moment see below
... img removed see orginal post ...

Your mistake is to believe that power_limit is an absolute power value. It's a ratio like in MSI afterburner : a value of 100 is equivalent to the max "official" TDP of your card.
Try with a value of 72 like the one you are using in MSI Afterburner and you shall be in line to compare.

Edit : didn't watched this API before nor using this miner, then saw there was a tdp method.
Did you try to read the value before applying a value (device.get.power_limit) ?
sr. member
Activity: 463
Merit: 250
Excavator is amazing, stable and the speeds are usually very competitive. It will not last long and it will be the best miner out there for most of the algos.

a litle correction
- Excavator it have potential to become amazing = true
- Stable = sorry negative on equihash algoritm is like play rusion rulet Wink now it work, now it crash for no particulary reason
- Speeds = dev yes they work to tune & make it competitive

but if miner is unstable & unrilable part with best miner out of discution, no offense

today test run "1.4.3 alpha", hoping with "bug fixes & improvements" equihash was fix & nop same problem what i say from few version agow

Code:
- Test Excavator Start
    [[2018-02-02 12:44:36.164770] [thread=0x000010d8] [info]]
- Miner Stall > Crash (i check log & see moment when speed was drop)
    [[2018-02-02 13:00:47.461761] [thread=0x00000a10] [info]]
- I stop manually miner seen him stall, coming back to pc, from what i was need to do in house
    [[2018-02-02 13:10:51.452607] [thread=0x000010d8] [info]]
    Shutting down
- Open / Start Excavator agen
- Crash agen
[[2018-02-02 13:53:57.508518] [thread=0x000011e0] [info]]

yup very stable ... but my be excavator he not like me after abandon him from 1.3.2a & switch to other miner, because speed competition Smiley if we can joke Smiley
cheers

I mean why would you even mine equihash with this miner? That is one of the algos that excavator is not competitive. Both DTSM or Bminer are much faster, even EWBF is faster Smiley
jr. member
Activity: 119
Merit: 3
because usual i like to find a fix to a problem, after reading more about https://github.com/nicehash/excavator/tree/master/api i was say hey let try to use "device.set.power_limit" setting my be i will make Excavator to be stable & work nice agen!!! so i set 86W limit (this is my target set already in Afterburner) to be sure it will never exceed that value & GPU Clock it will never goo more that 1860 Mhz. SURPRISE Excavator FAIL to obey / fallow his OWN fraking setting ending crashing GPU clock (2088 - 2216) => ending crashing himself !!! Way ?!?!? i will like to undestend but that's dev job to look in code & see way is out of control, way he can't fallow his own setting ... way can't be stable

Setting extra add in json:
Code:
{"time":5,"commands":[
{"id":1,"method":"device.set.power_limit","params":["0","86"]}
]},

i make a image to show working stable miner & unstable one on crash moment see below
member
Activity: 68
Merit: 10
The first release of the MinerPro for Nicehash is published. Very interested in your opinion about the application.

https://github.com/EvgeniyKorepov/MinerPro-for-NiceHash

Short demo video https://youtu.be/zN5rWmuU2mc

You must download and run the Excavator yourself on your farms. The application does not start the excavator, it controls local or remote instances of the miner.

Features
  • The calculation of profits is made taking into account the cost of electricity. Including different cost of electricity during the day.
  • Individual overclocking GPUs for each algorithm.
  • Fine-tuning the parameters of algorithms.
  • Ability to change parameters of the overclocking during operation.
  • Fast, less than one second, switch between algorithms.
  • Failover - switching to other locations when the stratum server fails.
  • Dev fee 0.8% (one minute in two hours).

jr. member
Activity: 119
Merit: 3
Excavator is amazing, stable and the speeds are usually very competitive. It will not last long and it will be the best miner out there for most of the algos.

a litle correction
- Excavator it have potential to become amazing = true
- Stable = sorry negative on equihash algoritm is like play rusion rulet Wink now it work, now it crash for no particulary reason
- Speeds = dev yes they work to tune & make it competitive

but if miner is unstable & unrilable part with best miner out of discution, no offense

today test run "1.4.3 alpha", hoping with "bug fixes & improvements" equihash was fix & nop same problem what i say from few version agow

Code:
- Test Excavator Start
    [[2018-02-02 12:44:36.164770] [thread=0x000010d8] [info]]
- Miner Stall > Crash (i check log & see moment when speed was drop)
    [[2018-02-02 13:00:47.461761] [thread=0x00000a10] [info]]
- I stop manually miner seen him stall, coming back to pc, from what i was need to do in house
    [[2018-02-02 13:10:51.452607] [thread=0x000010d8] [info]]
    Shutting down
- Open / Start Excavator agen
- Crash agen
[[2018-02-02 13:53:57.508518] [thread=0x000011e0] [info]]

yup very stable ... but my be excavator he not like me after abandon him from 1.3.2a & switch to other miner, because speed competition Smiley if we can joke Smiley
cheers
Pages:
Jump to: