Optiminer/Zcash v1.7 (GPU, Windows+Linux, AMD)

[tiennou]

it would be nice to know what kind of difference in hashrate is there before spending all that time and effort.

It substantial. Like ~15% based on what I've read for GCN 4th gen.

[1714065366]

1714065366

[wacko]

It substantial. Like ~15% based on what I've read for GCN 4th gen.

Thanks. I'm interested in the actual hashrates though, because there are other miners like the Claymore's that work in Windows. So, I'm getting 220 sol/s from RX 470/4, and curious whether installing and setting up a Linux system is worth it (if it's like 230 sol/s, then it's probably not, but can't decide without knowing the actual hashrates).

[nerdralph]

Version 1.4.0 Released!

https://github.com/Optiminer/OptiminerZcash#optiminerzzcash

[1.4.0] Experimental asm support for GCN 1 devices. Enable with '--experimental-kernel'
[1.4.0] 1-2% increase in hash speed.
[1.4.0] Fix race condition that could lead to invalid solutions.

Tried Tahiti & Pitcairn. Lots of "solution out of range" errors, even when I forced intensity 2.

[nerdralph]

Great work Optiminer  
is there anyone here have tonga to share with me hashrate
i got between 175 Sol/s and 199 Sol/s for each

I'm getting 200-210 with 1.4.0.  900/1500 clocks, using 1325 memory strap. Ubuntu 14.04/fglrx.

[dbct]

Just finished reading the whole topic, but still couldn't find what I was looking for. It's said that hashrate in Linux is faster for newer cards, but how much faster? What are the hashrates for RX 470 4GB and RX 480 8GB in Linux and Windows? For Linux noobs like myself it'll take the whole day to read all the guides out there and setup a Linux system so it would be nice to know what kind of difference in hashrate is there before spending all that time and effort.

hello wacko

we are working on what others have asked that relates to your question.
we are working on a hasrate page that will reflect good hashrate reports from people using our linux based mining OS
we have a raw page up at the moment but you can use it for comparison so you are able to diside if you want to go for windows or for a nice very easy to use linux based mining os (one of the miners that you can compare it great with is optiminer and we have there latest version incorporated)
the link to the raw sheet https://getpimp.org/pimps-corner/hashrates/

hope this helps you in your search for the perfect solution for yourself and your mining experience

Sincerely,
dbct
Community/Testing/Support Manager
Portable Instant Mining Platform (getPiMP.org)

[thilan29]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is my bat file:
optiminer.exe -s equihash.usa.nicehash.com:3357 -u someuser -p z -d 2 -i 6 --experimental-kernel

[johnkill]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is my bat file:
optiminer.exe -s equihash.usa.nicehash.com:3357 -u someuser -p z -d 2 -i 6 --experimental-kernel

the same with me for r9 290, and sapphire trix 390.

[cleverserge]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is my bat file:
optiminer.exe -s equihash.usa.nicehash.com:3357 -u someuser -p z -d 2 -i 6 --experimental-kernel

Try -i 5

[Holyblood89]

help me please. I sometimes miner sequence stops at the farms.
--watchdog i can not understand  <----sample can please

[johnkill]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is my bat file:
optiminer.exe -s equihash.usa.nicehash.com:3357 -u someuser -p z -d 2 -i 6 --experimental-kernel

Try -i 5

-i 5 ; -i 4 ; w/o -i the same error and 0 S/s
v 1.3 ... work great and I back to 1.3.2 - its still better than last Claymore 11.1

[thilan29]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is my bat file:
optiminer.exe -s equihash.usa.nicehash.com:3357 -u someuser -p z -d 2 -i 6 --experimental-kernel

Try -i 5

-i 5 ; -i 4 ; w/o -i the same error and 0 S/s
v 1.3 ... work great and I back to 1.3.2 - its still better than last Claymore 11.1

Yeah I also get 0S/s at any intensity.  The GPU is doing something though since the GPU usage shoots to 100%, so I'm not sure what is wrong.
v1.3.2 works fine for me.

[stoon]

v 1.4

Win 10, 15.12

6 x 290 = 1815 S/s @ 1300 watts 1.4
-75mv 900/1125 modded BIOS's

6 x Fury = 2010 S/s @ 748 watts 2.69
-96 800/400

1 x 290 + 5 x Nanos = 1995 @ 935 watts 2.13
-75 900/1125 + -96 800/400

Nice work.

[optiminer]

Just finished reading the whole topic, but still couldn't find what I was looking for. It's said that hashrate in Linux is faster for newer cards, but how much faster? What are the hashrates for RX 470 4GB and RX 480 8GB in Linux and Windows? For Linux noobs like myself it'll take the whole day to read all the guides out there and setup a Linux system so it would be nice to know what kind of difference in hashrate is there before spending all that time and effort.

It is not too hard to install ubuntu. And you just need a basic installation and download and run the miner from there.

If this is still too hard you can use something like ethos http://ethosdistro.com/

[optiminer]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is a problem with nicehash and v1.4. Use a different pool or run 1.3.2 on nicehash.

Will try to provide a fix.

The root of the problem is that nicehash sometimes uses extremely long extranonces . This violates some of the optimizations I made for v1.4.0 which causes this error (it should have printed a nice log message but there was a bug there which ended up in this cryptic message).

[charles2k]

Great work Optiminer  
is there anyone here have tonga to share with me hashrate
i got between 175 Sol/s and 199 Sol/s for each

My MSI R9 380 2G default clock (980/1380) only 158 S/s (ethos 1.1.8., generic kernel, with experimental kernel it has erros)

[jstefanop]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is a problem with nicehash and v1.4. Use a different pool or run 1.3.2 on nicehash.

Will try to provide a fix.

The root of the problem is that nicehash sometimes uses extremely long extranonces . This violates some of the optimizations I made for v1.4.0 which causes this error (it should have printed a nice log message but there was a bug there which ended up in this cryptic message).

Are you going to fix the index out of range errors with the experimental kernel on tahiti cards?

[thilan29]

I'm getting an "argument index out of range" error with v1.4 on a 290.  Works fine on 1.3.2.  Tried changing intensity values, taking out "experimental-kernel" but no luck.  Anyone else?

This is a problem with nicehash and v1.4. Use a different pool or run 1.3.2 on nicehash.

Will try to provide a fix.

The root of the problem is that nicehash sometimes uses extremely long extranonces . This violates some of the optimizations I made for v1.4.0 which causes this error (it should have printed a nice log message but there was a bug there which ended up in this cryptic message).

Thanks for the heads up.  I went back to 1.3.2.

[optiminer]

Are you going to fix the index out of range errors with the experimental kernel on tahiti cards?

AFAIK the index out of range errors only happen with nicehash, this does not depend on the kernel.

The experimental kernels seem generally to produce wrong solutions on the GCN1 cards like thaiti. That's why I called them experimental I need to get hold of such a card to debug which can take some days.

[cryptonitro]

Hi
i had gpu3 or gpu 5 stop working fast after launch , dont know how to target which one is it .... and reason

thx

[nerdralph]

Are you going to fix the index out of range errors with the experimental kernel on tahiti cards?

AFAIK the index out of range errors only happen with nicehash, this does not depend on the kernel.

The experimental kernels seem generally to produce wrong solutions on the GCN1 cards like thaiti. That's why I called them experimental I need to get hold of such a card to debug which can take some days.

I think you've made enough fees to buy a couple old used cards.  For my OpenCL test rigs I've go a few Pitcairn, a few Tonga, one Hawaii, one Tahiti, and one Ellesmere card.