• Forum has been upgraded, all links, images, etc are as they were. Please see Official Announcements for more information

X11 Mining Optimisation Project

Well you're still hashing at 96 megahash and hitting 2nd place on darkcointalk lol.
As he said he does it for supporting the nethash heathy / renting.... Ignite sorry for overtaking your replay ;)
 
Won't the optimizations for cudaminer be transferable as a piece of code into sgminer, or is it more tricky than that?
 
Won't the optimizations for cudaminer be transferable as a piece of code into sgminer, or is it more tricky than that?
That's the kind of thinking that got the current AMD X11 miner so shitty. You have to code to the strengths of your platform, not shoehorn code that's good on one platform onto another and make it run. CPU code rammed onto AMD GCN devices is pretty much darkcoin-mod in a nutshell.

Anyways, I booted into Windows to give my 290X cards more juice and see what they could REALLY do on X11 with my current code - was awesome, clocked one at 1200/1700. Screenshot (NSFW): https://ottrbutt.com/miner/x11localrigwolf-11262014.png
 
Wolf, OMG!! I feel so inadequate now :what: . What amount of ass licking and payoffs is it gonna take in order to get you to release a properly written and professional X11 miner to us noobs then? :rolleyes:
 
Wolf, OMG!! I feel so inadequate now :what: . What amount of ass licking and payoffs is it gonna take in order to get you to release a properly written and professional X11 miner to us noobs then? :rolleyes:

Me? Dude, I am probably not yet capable of properly writing X11 - I've done it decently, that's all. A properly written X11 miner wouldn't have all the hashes calculated by a single thread, Lord knows at least SIMD is horrible on GPU when done this way. While my kernels depend a lot less memory than the stock one, it still depends on memory because SIMD is so huge and poorly done that it accesses global memory a lot.

A PROPERLY done X11 would likely not even use the AES tables - the dev would bitslice the AES S-box, and that implementation would help out Groestl a lot, too, which is one of the slower ones, currently. SIMD would be split at least 4-way, probably 8-way, meaning that 8 threads are used to calculate a single hash. GPUs hate large amounts of code and complex tasks - they excel at tons of work in tiny pieces done by many different threads. BMW isn't as good as it could be, in stock or mine - Lord knows how to fix that, but I know there's gotta be a better way. Luffa, another kinda slow one, can easily be done in parallel - what stopped me is the ugliness of the SGMiner host code; I decided not to edit it, and confine my work to the OpenCL device code only. CubeHash isn't too slow, but a 2-way will probably make it faster.

In short, I am NOT who you're looking for - I am no professional, not yet.
 
Compiled a new binary from the github

http://www.filedropper.com/release12

In this version I have improved the hash for 970/980 cards.
This version supports Epsylon3's decimal intensity for optimal tweaking. -i 20.6 (etc)
It also contains some optimalizations from Schleicher.
 
This is with new #12 release.

-i 21

Thanks SP

Where can i donate?


ccminer.jpg
 
This is with new #12 release.
-i 21
Thanks SP
Where can i donate?
View attachment 665

My donation adresses:

BTC: 1CTiNJyoUmbdMRACtteRWXhGqtSETYd6Vd
DRK: XdgfWywdxABwMdrGUd2xseb6CYy1UKi9jX

you can try decimal intensities as well. f.ex -i 20.6 or -i 20.4 etc.. To find the optimal setting for your card
 
Coins coming your way.

If you can get this miner way up there i'll give some more.
 
Last edited by a moderator:
Coins coming your way.
If you can get this miner way up there i'll give some more.

Thanks :) I will continue to improve the hashrate, this is only the beginning... NVIDIA cards used to be used by gamers, and now the gamers can create Darkcoins.. And secure the blockchain...
 
Good news, this optimisation project is finally going ahead, I will be releasing final details in the next few days...
 
No one is discussing power usage and temperatures! I'd love to get some figures, I'm planning to get back into mining ;)
 
No one is discussing power usage and temperatures! I'd love to get some figures, I'm planning to get back into mining ;)
Nvidia is still on top with regards to efficiency...

Good timing, this optimisation project is going to be a big win for GPU Miners.
 
No one is discussing power usage and temperatures! I'd love to get some figures, I'm planning to get back into mining ;)

A standard 750ti card uses 40 watt of power when mining x11 in the latest miner. It doesn't need a 6pin powerconnector, and they stay cool.

This card is performing @ 2,7MHASH not overclocked, and 3,100 with overclocking.

http://www.gainward.com/main/vgapro.php?id=926&lang=en

For 6 cards a 500W psu is enough. But the 970 cards are probobly bether.
 
A standard 750ti card uses 40 watt of power when mining x11 in the latest miner. It doesn't need a 6pin powerconnector, and they stay cool.

This card is performing @ 2,7MHASH not overclocked, and 3,100 with overclocking.

http://www.gainward.com/main/vgapro.php?id=926&lang=en

For 6 cards a 500W psu is enough. But the 970 cards are probobly bether.

6 x 750ti GA Blacks pulled 400 Watts, that's with a HDD as well. They are fairly OC'd out of the box.

Standard 750ti's x 6 pulled around 340-360...

They were my experiences...
 
The 750ti GA black edition has a 6pin powerconnector. The bios on this card has probobly also been changed to support 65W.
The cards without the 6pin connector has a blocked bios that will prevent them to use more than 38-45 Watt. I am happy with the cheap gainward card,
but the black gives 200-300 KHASH more.
 
Hello everyone.

It's weird. I tested versions 11 and 14 but are not good results.

2 x 750ti shows 5,8MHASH in ccminer but the pool shows 1500 less (4,300MHASH).

With which drivers work OK?

Thanks in advance
 
Back
Top