Well, nanashi did the shuffle version, not trivial and almost 2x faster. Then I added another 20% over it :-D
Was this version ever published? )) 
Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 539. (Read 2347641 times)
Modern lyra2/lyra2v2 cuda miners are courtesy of nanashi ))
Nanashi reforked and added some percent.
Well, nanashi did the shuffle version, not trivial and almost 2x faster. Then I added another 20% over it :-D
Pallas, you are another big brain here, please, develop a new equihash cuda miner =D
Modern lyra2/lyra2v2 cuda miners are courtesy of nanashi ))
Nanashi reforked and added some percent.
Well, nanashi did the shuffle version, not trivial and almost 2x faster. Then I added another 20% over it :-D
nice job! LOL
by the way what is the current state of the miner,did you done more optimizations ?
I think we'll not have nvidia miner this week.. 
Genoil launched a new version of the amd miner, perhaps to be much more faster than the previous ones.
Pleaase, one good Nvidia miner!
Pleaase, one good Nvidia miner!
Guys, I'd advise not to mine zcash directly on nicehash's zcash pool. I was mining there and on coinmine at the same time since yesterday and my overall balance on nicehash is roughly 1/7th of what it should have been based on my coinmine earnings My hashrate is reported just fine on their pool and their miner is faster but still. And apparently their payouts doesn't work yet.
I can confirm - i'm meet the same shit
It's better to avoid NiceHash
when you do 30sols with a 280x.. you know the 1070 is misbehaving at 32 
Maximum hashrate with zcash on suprnova or coinmine with gtx 1070 now is 32 sol , with djm34 private miner up to 140 sol ? .
The amd genoil miner promise 100 sol for amd 390 card .
I have only nvidia card
.
Pls release a private miner for nvidia pascal , im happy to donate for have one .
The amd genoil miner promise 100 sol for amd 390 card .
I have only nvidia card
.Pls release a private miner for nvidia pascal , im happy to donate for have one .
Great work on the the Miner djm! very impressive mate 
looking forward the the crowd funding I will donate for sure..
you got any 750ti's to test with your code djm34?
looking forward the the crowd funding I will donate for sure..
you got any 750ti's to test with your code djm34?
He reported 55 s/s with a 780, you can expect 35-40 on a 750TI
Great work on the the Miner djm! very impressive mate
looking forward the the crowd funding I will donate for sure..
you got any 750ti's to test with your code djm34?
looking forward the the crowd funding I will donate for sure..
you got any 750ti's to test with your code djm34?
Modern lyra2/lyra2v2 cuda miners are courtesy of nanashi ))
Nanashi reforked and added some percent.
In the early days some pools didn't have a duplicate checker, so if you sent the same solutions 10 times in a row you got credited and 10 times the payout.
I expect djm34's code to be the real thing. Look at his great work in the lyra2/lyra2v2 and neoscrypt kernels.
Modern lyra2/lyra2v2 cuda miners are courtesy of nanashi )) I expect djm34's code to be the real thing. Look at his great work in the lyra2/lyra2v2 and neoscrypt kernels.
thinking of doing a crowd funding as usual...
new current speed 133sol/sec (yeah it changes fast
still highly unoptimized
new current speed 133sol/sec (yeah it changes fast
still highly unoptimized Damn that is fast.. I think I need to go back to the drawing board..
Do you have a screenshot please?
here we go... :
(the number of sol is consistent with what gets reported by suprnova)Is this on le gtx 1070?
How much power is the kernel using?
Hi DJ, do you know if the 470 gb and 480 gb will get similar results?
Thanks
I doubt you'll get that sort of hashrate. Currently the 980 are at 115Sol/s, but I tested a 780ti this afternoon (while the code was still running at 110sol/sec on the 1070, although at that time the 980 were running at 88sol/s) it was running at around 55sol/s. So expect even more degradation for much older card (in principle could be made compatible but I don't think it would be worthwhile, not at this stage of development).
I haven't tried yet on the 750ti
AMD rx470 /480 cards?
It seems his kernel does 20 sol per iteration instead of 2 ...
I wonder if this will make founs solutions "valid" or will be deemed as "hack"
I wonder if this will make founs solutions "valid" or will be deemed as "hack"
In the early days some pools didn't have a duplicate checker, so if you sent the same solutions 10 times in a row you got credited and 10 times the payout.
I expect djm34's code to be the real thing. Look at his great work in the lyra2/lyra2v2 and neoscrypt kernels.
Apparently this thread is attracting eth-babies. ~_~
GET OUT THERE AND EAT THE CODE DEVELOPER'S BRAINS!!!--
Then we would be fat and happy for a while. Happy Halloween! --scryptr
Then we would be fat and happy for a while. Happy Halloween! --scryptr
We should stop speculations and assumptions until we know more..
Its all on djm34 if, when and how he will do it..so we can just stay patient.
Its all on djm34 if, when and how he will do it..so we can just stay patient.
It seems his kernel does 20 sol per iteration instead of 2 ...
I wonder if this will make founs solutions "valid" or will be deemed as "hack"
I wonder if this will make founs solutions "valid" or will be deemed as "hack"
Jump to: