Scrypt has always been faster on x86 for me since I ever compiled cudaMiner.

oops, mining MRC with -C 2 using the Z kernel is not validating for me. Some problem
in the kernel it appears... Nothing to do with yesterday's commit, as I haven't touched
that kernel lately...

Strictly speaking the Z kernel shouldn't need support for any -C flags, as a Compute 3.5
device is capable of running any read access through the texture cache...

With -a=scrypt-jane:MRC -H 2 -C 0 -l Z60x24  I get up to 1100 kHash/s per GTX 780Ti.

EDIT: hmmm, two out of 3 cards hit 90 deg C and ran into thermal throttling.

Christian

erm apart from 2 solo mined blocks, and a 5.05 million MRC donation there was not
much of a return. I can call myself a Micro-Millionaire now at least.

I only started throwing massive power at a pool just now. Will check my returns in 24h.

Christian

hey, I think I just spotted a bathrobe on easy-mining.net

N=9

Lmao

One thing that sucks about N-factor changes is that the actual difficulty adjustment only takes place later (e.g in 615 blocks from now).. which means that the next 615 blocks will take 64 seconds to find - and not 32 as intended. Meh.

A good coin design could automatically downscale the difficulty at each N-factor change.

Also there is no good reason why the scrypt N parameter always has to grow in powers of 2 (apart from the tiny detail that in the code you can occasionally write &(N-1) instead of %N which is a tiny bit faster.

Again the same - lame question - can you upload please binary win x64 please? ;-)

does it matter how many connections you have with you wallet i have 4 on one comp 8 on the other

thanks does overclocking you card help