Author

Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 605. (Read 2347641 times)

legendary
Activity: 1470
Merit: 1114
where did you download cpuminer-opt/3.4.1?

See my sig.
legendary
Activity: 1764
Merit: 1024
Can't find a link to the developer enhanced GPU miner, apparently there was one made by Catia. Not sure if this Git is the same one or not https://github.com/MiniblockchainProject/CudaMiner
sr. member
Activity: 379
Merit: 250
where did you download cpuminer-opt/3.4.1?
legendary
Activity: 1176
Merit: 1015
Biggest mining profits are gone on XCN, I did some amd & nvidia on it but I must say that was the first time I missed my long time ago sold 750ti´s.

XCN did +600% and everyone in altmining section was talking about eth and lbry... it is so different from 2014 and 2015. There are gems for small miners almost every day, no easy walkthroughs available though. Currently again testing one old coin with no pools, looks like profitable.

As long as XCN is the only project with miniblockchain it just cannot die. Even if it gets delisted from every exchange there is. It is really strange that two years have passed and no one implemented this tech to bigger projects.
legendary
Activity: 2716
Merit: 1094
Black Belt Developer
Tried compiling ccminer m7, fixed some issues but still not OK.
Will work on it again tomorrow.
legendary
Activity: 1764
Merit: 1024
Sorry guys, but what are your calculations?

http://xcn.1gh.com/user?wallet=CJSVEgeeLYkPbJReCmPunDjKDgruqs6rq4

400MH/s

a 1070 if it's steady at 15, it's making 2.84 bucks/day, gross.

Yeah not very good, worse then ETH. If it's worse then ETH, it's not worth mining.

It probably could use a update if it hasn't been touched since 2015. Considering coinmarketcap just shows a huge pump and that's pretty much it, I don't think it's worth developers investing much time in... It could also be people who have been mining for prolonged periods of time are using a custom miner too...

There isn't even a block explorer anymore.
legendary
Activity: 3248
Merit: 1070
because it's not optimized at all, it still using old cuda compute, the code is from 2014
sr. member
Activity: 445
Merit: 255
why 30 if a 1070 hashes at 15 ?

Suprnova, in my experience, tells every time a different story...

Based on real 24h calculations, with 14/15 MH/s a 1070 can arrive at 2.84$, so a lot less profitable than lbry, right now and with the current available miners
legendary
Activity: 3248
Merit: 1070
Sorry guys, but what are your calculations?

http://xcn.1gh.com/user?wallet=CJSVEgeeLYkPbJReCmPunDjKDgruqs6rq4

400MH/s

a 1070 if it's steady at 15, it's making 2.84 bucks/day, gross.

maybe bad luck, suprnova is telling a different story, still at 30MH would be almost $6 per day which is more than lbry currently
sr. member
Activity: 445
Merit: 255
Sorry guys, but what are your calculations?

http://xcn.1gh.com/user?wallet=CJSVEgeeLYkPbJReCmPunDjKDgruqs6rq4

400MH/s

a 1070 if it's steady at 15, it's making 2.84 bucks/day, gross.
legendary
Activity: 1797
Merit: 1028

i don't think it work that way, because the 980ti can do only 14MH, but yes i also think a 1070 might do 20MH, but 40MH seems too much
I was able to get 15mh from gtx1070 with high oc on old miner. It is very bad. Good miner will do 30+ easily but again ... this coin is dead ... no need to bother.

CRYPTONITE IS TRADING AT 2X FORMER VALUE--

It has periodic spikes, and last week went to 4X-5X base market value.  The code just needs a touch-up to bring it into current standards, and to optimize it for Maxwell and Pascal.  I don't own any Pascal cards yet, too soon for me.  But my 980ti should be hitting 20MH/s or better.  I haven't run the numbers yet, it has been less than 24hrs, but XCN is earniing about the same as LBRY.       --scryptr
legendary
Activity: 3248
Merit: 1070

i don't think it work that way, because the 980ti can do only 14MH, but yes i also think a 1070 might do 20MH, but 40MH seems too much
I was able to get 15mh from gtx1070 with high oc on old miner. It is very bad. Good miner will do 30+ easily but again ... this coin is dead ... no need to bother.

why dead, if you can really do 30MH with one, this thing will be super profitable, one 1070 would do almost 0.02 per day
legendary
Activity: 1510
Merit: 1003

i don't think it work that way, because the 980ti can do only 14MH, but yes i also think a 1070 might do 20MH, but 40MH seems too much
I was able to get 15mh from gtx1070 with high oc on old miner. It is very bad. Good miner will do 30+ easily but again ... this coin is dead ... no need to bother.
legendary
Activity: 3248
Merit: 1070
there is definitely big profit for nvidia on m7 algo, if it can be added as a main algo, would be good

maybe we can rise a donation for this
legendary
Activity: 1797
Merit: 1028
Anyone know where djm34 is? Maybe hes using  the other me ?

DJM34 MUST BE QUIET--

Haven't seen him in a while.  His M7 branch CCminer (Cryptonite, XCN) could use some optimization.  It runs, but it hasn't been touched since early 2015.  It could be optimized for Maxwell, and Pascal, and given the API fetures and stuff.  Or, blended into Epsylon3's release.

Is everyone afraid to touch it?       --scryptr

criptonote for pascal is pretty bad at the moment, running at 350 for a 1070, when it should be like 800, clearly not optimized

Cryptonite (M7) is not cryptonight.

i mean cryptonight yes, bad typo, the one of monero, it's called also cryptonote i always confuse that...

CRYPTONITE (XCN)  USES M7 ALGO--

And the various currencies and algos are confusing.  I never bothered to mine XCN until I got questioned about it in another thread.  My 980ti gets 11.5MH/s with the CudaMiner XCN miner, and 14MH/s on DJM34's CCminer M7 branch.  Right now, I am trying to figure out whether it earns as much as Library Credits (LBC).  I am positive that optimization of M7 algo code for Maxwell and Pascal would significantly increase the hash rate.  Claymore's fee-based AMD miner reportedly gets 9MH/s on a 290X, haven't tried it yet.  He gets a 5% feee, and just updated the miner to v2.2 for Win10 and etc.

And, to add to the confusion, what is this M7M algo?  CPU only, I think...       --scryptr

ah yes m7 algo, i remember i mined that long time ago in 2014, i thought it was dead already, i need to check the hashrate with a 1070

m7m is for magi a cpu only coin it seems

p.s. do you have the release for this, i can not compile at the moment, i want to see how much hashrate one 1070 do with the m7 algo

ok i tried, not optimized at all, getting 12MH with 100-120w consumption



I TRIED TO COMPILE DJM34's CCMINER IN LINUX--

I also got a compile error.  I believe it was coded for CUDA 6.5, and I have a CUDA 7.5 environment.  The links for the compiled Windows versions work at 1GH.com, the CryptoMining-Blog links are dead.  I would appreciate any coder's advice in how to get a Linux compile from the source at DJM34's github.  "It should compile with any version of CUDA 6.5+", was posted somewhere.  The source there does not include a "build.sh" file, and the code must be compiled with "./autogen.sh", followed by "./configure", and finally, "make".  I'd appreciate a "build.sh" that gets the job done.

CryptoNite (XCN) just had a major value spike.  Miners and traders are still interested.       --scryptr
legendary
Activity: 3248
Merit: 1070
if it was optimized for a 1070 it would be very profitable, should be at least 15MH per card, so the profit would be like lbry at least
gtx750 can do > 4mh. gtx1070 roughly 5-10 times faster then gtx750 so the good speed should be 20-40 ))

i don't think it work that way, because the 980ti can do only 14MH, but yes i also think a 1070 might do 20MH, but 40MH seems too much
legendary
Activity: 1510
Merit: 1003
if it was optimized for a 1070 it would be very profitable, should be at least 15MH per card, so the profit would be like lbry at least
gtx750 can do > 4mh. gtx1070 roughly 5-10 times faster then gtx750 so the good speed should be 20-40 ))
legendary
Activity: 3248
Merit: 1070
if it was optimized for a 1070 it would be very profitable, should be at least 15MH per card, so the profit would be like lbry at least
legendary
Activity: 1510
Merit: 1003
it was a short pump for xcn on polo, it was rather good to mine it with gtx750 using lowmem ccminer by djm. Now xcn is dead again ))
legendary
Activity: 3248
Merit: 1070
Anyone know where djm34 is? Maybe hes using  the other me ?

DJM34 MUST BE QUIET--

Haven't seen him in a while.  His M7 branch CCminer (Cryptonite, XCN) could use some optimization.  It runs, but it hasn't been touched since early 2015.  It could be optimized for Maxwell, and Pascal, and given the API fetures and stuff.  Or, blended into Epsylon3's release.

Is everyone afraid to touch it?       --scryptr

criptonote for pascal is pretty bad at the moment, running at 350 for a 1070, when it should be like 800, clearly not optimized

Cryptonite (M7) is not cryptonight.

i mean cryptonight yes, bad typo, the one of monero, it's called also cryptonote i always confuse that...

CRYPTONITE (XCN)  USES M7 ALGO--

And the various currencies and algos are confusing.  I never bothered to mine XCN until I got questioned about it in another thread.  My 980ti gets 11.5MH/s with the CudaMiner XCN miner, and 14MH/s on DJM34's CCminer M7 branch.  Right now, I am trying to figure out whether it earns as much as Library Credits (LBC).  I am positive that optimization of M7 algo code for Maxwell and Pascal would significantly increase the hash rate.  Claymore's fee-based AMD miner reportedly gets 9MH/s on a 290X, haven't tried it yet.  He gets a 5% feee, and just updated the miner to v2.2 for Win10 and etc.

And, to add to the confusion, what is this M7M algo?  CPU only, I think...       --scryptr

ah yes m7 algo, i remember i mined that long time ago in 2014, i thought it was dead already, i need to check the hashrate with a 1070

m7m is for magi a cpu only coin it seems

p.s. do you have the release for this, i can not compile at the moment, i want to see how much hashrate one 1070 do with the m7 algo

ok i tried, not optimized at all, getting 12MH with 100-120w consumption...also very unstable, bottom 50MH+ top was 72MH

Jump to: