I might do NeoScrypt for CUDA if noone else does, but probobly not.
NeoScrypt would also be nice, just forgot about it
Keeping my fingers crossed anyways
Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 1220. (Read 2347601 times)
NeoScrypt would also be nice, just forgot about it
Keeping my fingers crossed anyways
I might do NeoScrypt for CUDA if noone else does, but probobly not.
I got 2 DRK's Yesterday from someone. Thanks
in 5 years they might be worth $1000 each.
in 5 years they might be worth $1000 each.
I sure would donate, but I'm not really into X11-15. Have you considering having a look at CryptoNote/Monero or the just released SpreadCoin GPU miner (tsiv gave nvidia a nice advantage but amd is catching up)?
Nevertheless, keep up the great work
I got 2 DRK's Yesterday from someone. Thanks
in 5 years they might be worth $1000 each.
in 5 years they might be worth $1000 each.
You did say you would stop publishing when you got to 10MH/s on 980, remember?
After I went opensource, I have got 0.01 BTC in donation from one miner. I have released a miner that is 30% faster on the 980, and get nothing in return. Come on guys, Donate some Beers at least ?
I picked up a couple of 280x now from a tired miner. Got them for $150 a piece. Time to move on to OpenCL.
Why dont you just work together?
Pool your knowledge on github and make the best miner ever.
Why must it be like good luck, I know something you dont
Pool your knowledge on github and make the best miner ever.
Why must it be like good luck, I know something you dont
Intensity 21 slows down the pc. Anything higher crashes. I like 19 or 20 for the gtx970
Did you try release 11 and no intensity parameter? the default is set between 19 and 20 wich seems to be the fastest on 750ti. Not sure on the 970 though
After I went opensource, I have got 0.01 BTC in donation from one miner. I have released a miner that is 40% faster on the 980, and get nothing in return. Come on guys, Donate some Beers at least ?
I picked up a couple of 280x now from a tired miner. Got them for $150 a piece. Time to move on to OpenCL.
Have fun. Soon you'll learn that you won't get much of anything...I picked up a couple of 280x now from a tired miner. Got them for $150 a piece. Time to move on to OpenCL.
The problem is that the current prices are really low. A miner needs to get profit after the electricity bill is payed.
Anyway, my AMD miner will be private and closed source. Good luck with your sourcecode sale btw. Did someone really offer you 22.5 BTC for it?
Intensity 21 slows down the pc. Anything higher crashes. I like 19 or 20 for the gtx970
You did say you would stop publishing when you got to 10MH/s on 980, remember?
After I went opensource, I have got 0.01 BTC in donation from one miner. I have released a miner that is 30% faster on the 980, and get nothing in return. Come on guys, Donate some Beers at least ?
I picked up a couple of 280x now from a tired miner. Got them for $150 a piece. Time to move on to OpenCL.
result for xxxxx does not validate CPU is same as HW: in cgminer/sgminer.
If you get these you should lower the intensity or the clock, because they will lower the hashrate on the pool. You should also make sure that you are ruinning the latest NVIDIA driver (.75).
On the 970 the ccminer52.exe shoud be faster, however if you recieve alot of (does not validate) You can try the ccminer.exe.
If you get these you should lower the intensity or the clock, because they will lower the hashrate on the pool. You should also make sure that you are ruinning the latest NVIDIA driver (.75).
On the 970 the ccminer52.exe shoud be faster, however if you recieve alot of (does not validate) You can try the ccminer.exe.
Buildt a new version with tsivs x13 improvements.
Some bugfixes, and bether default intensity for 750ti (I get +30 KHASH)
version 10 had a memory leak. If the miner was running for a long time, the performance went down.
The 11th public beta(windows executable) is available here: (11-25-2014)
...
Win8.1 , driver 344.75, launching with with -i 20Some bugfixes, and bether default intensity for 750ti (I get +30 KHASH)
version 10 had a memory leak. If the miner was running for a long time, the performance went down.
The 11th public beta(windows executable) is available here: (11-25-2014)
...
EVGA GTX 970 SC ACX2 stock
X11 7320
X13 5900
X15 5180
NIST5 22650
I tried earlier today the link at the crypto mining blog
(http://cryptomining-blog.com/3899-updated-ccminer-1-5-git-fork-by-sp-optimized-for-maxwell/)
but it crashed after 20-30min (maybe less). I've been running release11 for 30min and NIST is varing (22400 as of now).
Also, what does this mean?
CPU #1: result for xxxxx does not validate CPU
or
CPU #1: result for xxxxx is not in range ......
I get these sometimes at X algos (more often the second one)
Is this "normal" or it's a big deal?
Impressive work!
Buildt a new version with tsivs x13 improvements.
Some bugfixes, and bether default intensity for 750ti (I get +30 KHASH)
version 10 had a memory leak. If the miner was running for a long time, the performance went down.
The 11th public beta(windows executable) is available here: (11-25-2014)
http://www.filedropper.com/release11
The sourcecode is available here:
https://github.com/sp-hash/ccminer
Some bugfixes, and bether default intensity for 750ti (I get +30 KHASH)
version 10 had a memory leak. If the miner was running for a long time, the performance went down.
The 11th public beta(windows executable) is available here: (11-25-2014)
http://www.filedropper.com/release11
The sourcecode is available here:
https://github.com/sp-hash/ccminer
yes, this is how the X-intensity is implemented in sgminer. Just Multiply the factor by 256.
-xi 128 = gives throughput of 128*256
or perhaps the constant is more than 256. Don't remember. But the idea is to multiply the constant with the inputparameter.
-xi 128 = gives throughput of 128*256
or perhaps the constant is more than 256. Don't remember. But the idea is to multiply the constant with the inputparameter.
I reset the default throughput to be the same as in build 8, it increases the x11 hash with around 30KHASH (750ti). Now trying out 72 registers and no launchbound as suggested @x13. But so slow to build...
I why i prefer to do tests on linux when i can... I work over nfs on windows and build in putty
Increasing too much intensity will make Windows very slow, its why i dont use the max possible as default. But i understand your ".5" intensity, could be done like that rounded to the nearest multiple of 256
for the api its normal... "-b 0" to fully disable it, or set a custom port if required... There is only 4 tries to bind the first minute, then stops. Nice also to know there is another instance running
I reset the default throughput to be the same as in build 8, it increases the x11 hash with around 30KHASH (750ti). Now trying out 72 registers and no launchbound as suggested @x13. But so slow to build...
Eplyson, the API seems to fail when you run two instances of ccminer. Probobly because the default port is used.
http://pastebin.com/sV7C2TWR
Try that for Hamsi. It sure is not pretty but it's a whole lot faster than the one I originally left X13 with. My notes say +59% on the 750 Ti
Yep, i get much more with your code Try that for Hamsi. It sure is not pretty but it's a whole lot faster than the one I originally left X13 with. My notes say +59% on the 750 Ti
i keep it tx.2250KHs on a 750 Ti, vs 2150 before changes, and slower with sp commit (tried differents maxregcount on linux)
edit: i used 72 registers, 64 is too low
I use launchbound 256,4 and only 32 threads. Seems to be the fastest on windows. I get no memory spills with 64 registers for compute 5.0, but for compute 3.0 there are spills (because of the missing funnelshift)
The best setting for the TI is betweeen 19 and 20, so I changed the default. Preparing a new build now. #11.
We should add xintensity to ccminer for optimal performance.
We should add xintensity to ccminer for optimal performance.
Jump to: