excavator by NiceHash - multi-algorithm NVIDIA GPU miner [1.5.6a]

[sp_]

Haven't seen a Pascal card to die because of high overclock. We did destroy one due to water leakage and some stuff was burnt on it, but never due to high memory clock and we are running them 24/7 mining. Just make sure you have adequate cooling, which you already need if you are mining 24/7.

I use oc and reduced TDP and have no 1070 cards RMA'd since september. (6 months running 100% mining)
Of the gtx 1060 3gb's I have RMA'd 1% (6 months running 100% mining)

[sp_]

This is intriguing when you run 8 rigs in an apartment (48x GTX 1070) as the outside temp starts to warm up.  It's also interesting because at 70-75 TDP I could easily run 3x rigs per 20A circuit. 

I run 3x gtx 1070 rigs on 16A without any problems.

[sp_]

I wrote in the other thread:

A possible speedup for equihash

Could the blake2s pass be removed completely (round 0)?  Dual mined  with the memory accesses? I am not talking about 2 threads, but one thread with the round0 merged into the other rounds.

The miner will need to do 1 nonce search(round1-round8) and one blake2s  (round0) merged into the round1-round8 code per iteration. Since you work on the round0 data from the previos run, the nonce found would be the result of the previous padding data.  

On NVIDIA round0 take 20% of the time. (The opensource ZEC DJezo kernel)

[giagge]

What this the .bat for mining zec on another pool with nvidia ? .

[djeZo]

I wrote in the other thread:

A possible speedup for equihash

Could the blake2s pass be removed completely (round 0)?  Dual mined  with the memory accesses? I am not talking about 2 threads, but one thread with the round0 merged into the other rounds.

The miner will need to do 1 nonce search(round1-round8) and one blake2s  (round0) merged into the round1-round8 code per iteration. Since you work on the round0 data from the previos run, the nonce found would be the result of the previous padding data.  

On NVIDIA round0 take 20% of the time. (The opensource ZEC DJezo kernel)

1 nonce is already used now, but each blake2s hash has different id, so you get out ~2 mil blake2s hashes. You could minimize read and writes, but then you would have to recalculate blake2s hashes each round (computation-memory tradeoff), which for sure would, even if you reached higher speed, cause much higher power consumption.

[Slava_K]

Can you read hashrate directly from miners and send to pool? This can make users hash rate more stable.

[Kayfolom]

Launch parameters for 5x1070
excavator.exe -a equihash -s equihash.eu.nicehash.com:3357 -u 1PJQRnBRZosg********.***07:x -ca -ca -or -os 0 100 700 60 -os 1 100 700 60 -os 2 100 700 60 -os 3 100 700 60 -os 4 100 700 60

Add to the description, it was necessary to find out experimentally

The results are not impressive. (At EWBF's  speed 2450H/s)

Code:

[20:52:28][0x00001794][info] net | Share #66 accepted
[20:52:29][0x00000c94][info] =====> 1.996253 kH/s
[20:52:29][0x00000c94][info] gpuinfo | GPU#0    TEMP: 50 C      USAGE: 99 %     MEMCTRL: 63 %   POWER: 106.192 W
[20:52:29][0x00000c94][info] gpuinfo | GPU#1    TEMP: 59 C      USAGE: 99 %     MEMCTRL: 75 %   POWER: 106.999 W
[20:52:29][0x00000c94][info] gpuinfo | GPU#2    TEMP: 59 C      USAGE: 99 %     MEMCTRL: 74 %   POWER: 106.553 W
[20:52:29][0x00000c94][info] gpuinfo | GPU#3    TEMP: 60 C      USAGE: 99 %     MEMCTRL: 76 %   POWER: 106.988 W
[20:52:29][0x00000c94][info] gpuinfo | GPU#4    TEMP: 60 C      USAGE: 99 %     MEMCTRL: 75 %   POWER: 107.099 W
[20:52:29][0x00000c94][info] =====> Total power: 533.831 W      Efficiency: 3.739485 H/W
[20:52:35][0x00001794][info] net | Share #67 accepted

[NiceHashSupport]

You are running cards with only 106 W power, no wonder you get low hashpower (400 per card). Now just increase TDP if you want more speed, eg set TDP to 70 and see how that improves hashrate.

EDIT: You did not show your memory clocks, but if you have samsung, you can easily go 4700. If you have micron, you can go 4400 stable and experiment with 4450.

[ioglnx]

Does excavator overwrites the oc setting by default I set with MSI Afterburner after I started excavator?
Or the settings are remain..because I set the oc after starting the miner with msi afterburner but I noticed it's still slower than EWBF by 20%.

[Amph]

i see that nexus is using sha3, is it the same as shat? isn't shat three times sha or there is something more?

nevermind the rounds are different...

[NiceHashSupport]

Does excavator overwrites the oc setting by default I set with MSI Afterburner after I started excavator?
Or the settings are remain..because I set the oc after starting the miner with msi afterburner but I noticed it's still slower than EWBF by 20%.

It applies overclocks x seconds after startup (by default, x is 2 seconds), so, it overwrites your previous overclocks set by any other tool. If you use another tool to override overclocks, excavator will not mess with that. If you used -or parameter, then clocks are reset to default 0 0 100 at exit of excavator.

[ioglnx]

I don't want excavatgor woverride my settings i apply after i started the miner.
I want to use my tool..how to achieve this? so it should take the setting i made with msi afterburner etc.. how to achieve this?

[NiceHashSupport]

I don't want excavatgor woverride my settings i apply after i started the miner.
I want to use my tool..how to achieve this? so it should take the setting i made with msi afterburner etc.. how to achieve this?

Then you don't set any overclocking in excavator. It is only optional, it can help you reach higher stable clocks because there is no switching to p0 state and it is free for anyone to use or not.

[Newton90]

have 85 - 87 hash per thread on asus gtx1060 o3g (65% tdp,core +170),total 255-261 hash.

on the same sets have 26x - 27x hash/s with ewbf.

without fee ewbf speed is faster.need improvements.

even more: new ewbf version 0.3.3 has --fee option with 0% fee set.so we need serious speed improvement.

[thhjs]

Some speed results:

equihash:
optimized GTX 1080 at 60% TDP = 500 Sol/s
optimized GTX 1070 at 60% TDP = 440 Sol/s

How do you get 440 Sol/s with 60 TDP ?

I'm getting 365.

My EVGA GTX 1070 SC is running at 60 % TDP + 200 Mhz Mem.

[NiceHashSupport]

Some speed results:

equihash:
optimized GTX 1080 at 60% TDP = 500 Sol/s
optimized GTX 1070 at 60% TDP = 440 Sol/s

How do you get 440 Sol/s with 60 TDP ?

I'm getting 365.

My EVGA GTX 1070 SC is running at 60 % TDP + 200 Mhz Mem.

i'm not making at all the 440 reported here, but only 1650(412 each) with 4 gpu, 60% tdp +100 core +600 ram

Make sure you are using at least 2 threads. If only 1, then your GPU is not 100% loaded. 1 thread solution is good if you plan to do other stuff with your card. Imagine it as being an intensity switch.

We reached 440 with 115W TDP limited card that has OC set to 2120 core and 4700 mem. With next version, it is about 445 with these same settings.

Another example is 126W TDP limited card, OCed to 2110 core and 4400 mem (micron with BIOS patch) reaching 440 sol.

You can view how far is your card OCed when you launch excavator with overclocking set.

[thhjs]

Missed that post sorry.
I will try it thanks.

[kopija]

Greetings,
out of curiosity, this miner is open to any pool and does not contain any fees, so it basically works against your core business of selling hashing power/providing pools.
So whats the rationale behind it?

[NiceHashSupport]

New version: http://nicehash.com/tools/excavator_v1.1.4a_Win64.zip

Changes:

- slight equihash efficiency improvement for GTX 1060 cards
- equihash GTX 750 Ti fix
- added missing MSVC files
- fixed issue of inability to display proper power usage
- general improvements

GTX 1060 6G gets 310 sol/s @ 90 W. GTX 1060 3G gets 255 sol/s @ 72 W.

[ExEric3]

New version: http://nicehash.com/tools/excavator_v1.1.4a_Win64.zip

Changes:

- slight equihash efficiency improvement for GTX 1060 cards
- equihash GTX 750 Ti fix
- added missing MSVC files
- fixed issue of inability to display proper power usage
- general improvements

GTX 1060 6G gets 310 sol/s @ 90 W. GTX 1060 3G gets 255 sol/s @ 72 W.

Hello,

nice release thanks.

I have some suggestion for next version:
- add FAN control for devices (so are not needed external utilities anymore)
- change -or function in overclocking section when you close miner will set default parameters (to 0 0 100) and I want them customisable for example I want set 60% TDP and prepare card for another algo and mining software. So again here eliminate some external utilities.

Its possible? Thanks for answer.