Topic: excavator by NiceHash - multi-algorithm NVIDIA GPU miner [1.5.6a] - page 44. (Read 179268 times)

Check your clocks (core and memory), and run with two or more workers.

I got an ASUS 1070 Turbo 8gb that is giving me trouble and I would like to know if maybe somebody else encountered this problem.

I have the card installed in windows 10 with the most recent drivers, and it is recognized by the system and by Afterburner, and also by excavator and nicehash miner.

The problem is - at stoc settings I am only getting 330 sols/sec. But no matter what OC settings I throw at it , it still remains at 330-340 (or even less if TPD is 70-75).

Any ideeas?

How can I use it with suprnova?

for 1070 equihash is this still the best miner? Haven't had time to spend on the forums lately but been running excavator for the last couple months, just checking if anything faster came out.

While it seems that Optiminer is still the best for RX480 cards with 305 in average in equihash.
Will wait for linux version to give it a try.

Thanks for the detailed feedback. The low hashrate for Equihash is most likely a driver issue.
(I am pretty sure I mentioned somewhere that you need Crimson 16.9.2 for optimal performance.)
The current developmental version runs at around 290 S/s on stock RX480 with the latest driver.
We will see if we can release it sooner than later along with Ethash support.

I checked excavator 1.2.4a (latest version 04.07.2017) speeds.

at the moment it only support 4 algos:
equihash
pascal
sia
decred

sia(nanopool) and decred(suprnova) wont work for me because of some problems with pool get job routines (failed to parse JSON from pool).

pascal gave ~710mh for RX 470 it is good, because whattomine show 560mh for RX 470.   RX 470+RX480+RX480 gave me 2gh at pascal.

equihash (ZCash) speed is only ~210,  claymore miner give me ~290mh. I setup 4 workers for each device (as manual said), tryed to switch on binary options and global data share (GDS wont work for rx 480, only rx 470) but it not helped much.

So for now excavator for AMD cards is bad at equihash, and good for pascal.

excavator have no dev fee but if compare to claymores 2% fee, claymore miner speed is much faster 290 sol - 2% = 284 sol and for now claymore is better.

Due to good structure, excavator allow to do any compilations of algos, so maybe some interest dual or triple mining schemes may result in good profits. Need to do research it that way.

For now it is not worth to switch from claymore miner.
Will wait ethash support.

Well, I would say you do the math. 

You mean faster than genoil , not claymore right?

Hard to say as I am quite ambitious with Ethash. We got a permission from Genoil to use his kernel, and it's already running faster by 15% after a few optimizations on RX 470. We will see.

Please see my reply to your PM. Did you try to overclock?
I ruled out that possibility because GPU-Z showed stock clocks.

I had a similar error with 1070 on another miner. Happens the card did not support memory overclocking.

hi everyone.

Got en error with excavator. i got a few cards, but Only this card 1070 Gygabyte gives this error. it is new. I wonder it is warranty, but not sure how to prove it.
if you know what to do please let me know please. the error shows up everytime i run the excavator and it does not let me do benchmark in nicehash.

the error as follows: Cude ........ 2276 , 2191

https://www.upload.ee/image/7074177/11.jpg
https://www.upload.ee/image/7074206/2017-06-02_013858.jpg

Any eta on ethash?

Yes, currently manual for this is being created, but drafted version is here: https://github.com/nicehash/excavator/blob/master/excavator-man-API.txt

It works fine for NVIDIA, regarding AMD, please ask zawawa.

I am currently working on it along with Ethash. No worries!

is it possible to set fan speed through "device.set"? If yes, is it in percentage? How would it look in json such a line?
It would be great to not have to use the afterburner for fanspeeds.
Or is there another alternative for fan speeds?

Are there any news about *nix version?

Morning.

Just started mining yesterday. Super excited

Started with EQM which is cool but I think I'd like to just get paid in ZEC.

Pointed at zecmine.pro and seems to be working but I wonder if anyone can clarify and comfort me that I'm getting full efficiency out of this.

I'm running two 1080ti cards BTX.

On the log it only shows one device:
[08:09:12][0x000004e4][info] net | Share #82 accepted
[08:09:16][0x000008d8][info] net | Share #83 accepted
[08:09:18][0x000004e4][info] core | Device #0-0 speed: 292.836667 H/s
[08:09:18][0x000004e4][info] core | Device #0-1 speed: 291.620000 H/s
[08:09:18][0x000004e4][info] core | Algorithm 'equihash' total speed: 1.225017 kH/s
[08:09:20][0x000008d8][info] net | Share #84 accepted
[08:09:21][0x000004e4][info] net | Share #85 accepted
[08:09:21][0x000008d8][info] net | Share #86 accepted
[08:09:27][0x000004e4][info] algo-equihash | New job '446626875279046', diff=481.882

Although the "total speed" is essentially the total of all four workers (two per GPU).

Also on the mining site it only shows one worker but hashrate of 1.25 KS/s.

So I guess just trying to wrap my brain around a worker count.

From the above does it look like I'm using my two GPUs at full potential? If so, great. If not I'd love some tips on how to tweak them.

Thanks.


EDIT: realized that pool had not found a block in 14 days so I switched to nanopool. Still only seeing device "0" but getting around 1.2 KH/s.

https://github.com/nicehash/excavator/releases