Anyone know what x11gost algo is? SP ? Is that in ccminer ? thx
SIbcoin. Try tpruvot's.
Topic: CCminer(SP-MOD) Modded NVIDIA Maxwell / Pascal kernels. - page 715. (Read 2347641 times)
SIbcoin. Try tpruvot's.
Slower hashrates on three I tried. ~15% less.
I have tested this now. Slower on the 750ti. The new blake implementation works good on compute 5.2, but on the 750ti 10% slower..
is this with faster vcash and can mine decred?
Not faster. For a faster Vcash kernal. donate 0.2BTc. (15% faster than the Provos Alexis kernal) And get decred #7 included. (25-30% faster than 1.7.5)I get
3150 on 1450
3300 on 1466
with decred private #4
Anyone know what x11gost algo is? SP ? Is that in ccminer ? thx
SIbcoin. Try tpruvot's.
Anyone know what x11gost algo is? SP ? Is that in ccminer ? thx
i tried 1.579 but it is slower than the djrm34 mod for neoscrypt a bit slowe
also more not validate nonce, djm34 version is better
try quark.also more not validate nonce, djm34 version is better
Here is what I get on the standard clocks: 1290Mhz core. release 79 is around 500khash faster than 78 on the 980ti windforce G1 gaming
release 79 crashed mining quark on 2gb compute 5.2 cards. (gtx 950,960) I submitted a fix
release79.7z updated on github with new exe.
release79.7z updated on github with new exe.
is this with faster vcash and can mine decred?
Not faster. For a faster Vcash kernal. donate 0.2BTc. (15% faster than the Provos Alexis kernal) And get decred #7 included. (25-30% faster than 1.7.5)
i can get 3000 mega on xvc with 1.7.4(1450 core) is your version faster?
i tried 1.579 but it is slower than the djrm34 mod for neoscrypt a bit slowe
also more not validate nonce, djm34 version is better
try quark.also more not validate nonce, djm34 version is better
Here is what I get on the standard clocks: 1290Mhz core. release 79 is around 500khash faster than 78 on the 980ti windforce G1 gaming
i tried 1.579 but it is slower than the djrm34 mod for neoscrypt a bit slowe
also more not validate nonce, djm34 version is better
also more not validate nonce, djm34 version is better
try quark.
1.5.79(sp-MOD)
-Faster lyra2v2
-Faster quark
-Faster neoscrypt
-Faster x11,x13,x14,x15,x17,nist5
-Faster deep and doom
-Added vcash (blake 256 (run with -a vanilla))
1.5.79(sp-MOD) is available here: (02-april-2015)
https://github.com/sp-hash/ccminer/releases/
The sourcecode is available here:
https://github.com/sp-hash/ccminer
Please test it and report your hashrates. Compare with 1.5.78. What are you getting?
Slower hashrates on three I tried. ~15% less. -Faster lyra2v2
-Faster quark
-Faster neoscrypt
-Faster x11,x13,x14,x15,x17,nist5
-Faster deep and doom
-Added vcash (blake 256 (run with -a vanilla))
1.5.79(sp-MOD) is available here: (02-april-2015)
https://github.com/sp-hash/ccminer/releases/
The sourcecode is available here:
https://github.com/sp-hash/ccminer
Please test it and report your hashrates. Compare with 1.5.78. What are you getting?
Hrmm at this point I cannot remember the title but I remember it said tpruvot 1.7 or so. Obviously a scam since when after the fact I went to the github and saw tpruvot wasn't even on 1.7 - lol - hind sight is always 20/20.
Any chance you mined Oxen?
Some people confused ccminer having malicious code instead of Oxen which was malicious.
cryptomining blog really? that site is huge for mining, how they can not check for malicious code, are they stupid? what was the miner with the virus?
Apparently people can hide their malicious stuff with FUD crypters so the virus can't be detected and can be triggered anytime, even like months later.
it's not about antivirus, it's about code checking, no virus can be hidden if you have the source and you know what you're doing
They obviously publish a clean surce with dirty binaries from a the private, malicious source. People would be fine compiling the clean source but they rarely do it themselves on windows.
soon : http://techcrunch.com/2016/03/30/be-very-afraid-hell-has-frozen-over-bash-is-coming-to-windows-10/
That will be very interesting. It might even convince me to make the switch from 7 to 10.
Someone told me he is ... can't say.
I read somewhere that you got "robbed" by someone copying your wallet dats and actually got into your system with teamviewer?!
There are a few people who got robbed that way (search reddit).
It actually seems teamviewer is the culprit here.
Imo someone is bruteforcing teamviewer accounts. The attacker only needs to bypass the loginlimit somehow ...
Don't forget the basic teamviewer password is only 6 letters, so even if you set a permanent login password that is way longer that basic 6 letter pw is still valid.
(the default 6 letter pw can be changed to a 10 letter pw including special characters)
There are a few people who got robbed that way (search reddit).
It actually seems teamviewer is the culprit here.
Imo someone is bruteforcing teamviewer accounts. The attacker only needs to bypass the loginlimit somehow ...
Don't forget the basic teamviewer password is only 6 letters, so even if you set a permanent login password that is way longer that basic 6 letter pw is still valid.
(the default 6 letter pw can be changed to a 10 letter pw including special characters)
I doubt it's Teamviewer because it increases the amount of time you have to wait to try again after unsuccessful tries but it's possible.
But, you can change the default password's length to 10 characters which includes symbols in Options\Security > password strength.
And you can also disable incoming connections from outside your local network. It also makes windows much more responsive and don't use your bandwidth and of course it's completely safe from the outside (Options\General - Incoming LAN connections > accept exclusively). In this case you can also use local IP addresses to connect for example in shortcuts like this which is pretty convenient:
"C:\Program Files (x86)\TeamViewer\TeamViewer.exe" -i 192.168.1.x --Password doesnt_matter_local_only_password
The path is to your local teamviewer, the example above is what you'd be putting in a shortcut link to easily connect to a local PC.
And as far as I know IP addresses instead of IDs are only allowed for LAN connections.
You still should have a password set on the remote PC it just doesn't have to be strong if you set it that it only accepts connections from inside your local network.
I read somewhere that you got "robbed" by someone copying your wallet dats and actually got into your system with teamviewer?!
There are a few people who got robbed that way (search reddit).
It actually seems teamviewer is the culprit here.
Imo someone is bruteforcing teamviewer accounts. The attacker only needs to bypass the loginlimit somehow ...
Don't forget the basic teamviewer password is only 6 letters, so even if you set a permanent login password that is way longer that basic 6 letter pw is still valid.
(the default 6 letter pw can be changed to a 10 letter pw including special characters)
There are a few people who got robbed that way (search reddit).
It actually seems teamviewer is the culprit here.
Imo someone is bruteforcing teamviewer accounts. The attacker only needs to bypass the loginlimit somehow ...
Don't forget the basic teamviewer password is only 6 letters, so even if you set a permanent login password that is way longer that basic 6 letter pw is still valid.
(the default 6 letter pw can be changed to a 10 letter pw including special characters)
I doubt it's Teamviewer because it increases the amount of time you have to wait to try again after unsuccessful tries but it's possible.
But, you can change the default password's length to 10 characters which includes symbols in Options\Security > password strength.
And you can also disable incoming connections from outside your local network. It also makes windows much more responsive and don't use your bandwidth and of course it's completely safe from the outside (Options\General - Incoming LAN connections > accept exclusively). In this case you can also use local IP addresses to connect for example in shortcuts like this which is pretty convenient:
"C:\Program Files (x86)\TeamViewer\TeamViewer.exe" -i 192.168.1.x --Password doesnt_matter_local_only_password
I am working on the sp-mod ultimate. Most of the kernals merged into one. Since #6 will include decred,vcach,blake2s the price will be more than 0.1 BTC.
I have added some more quark hash:
This private miner will earn 0,00142592BTC / day @32MHASH quark. (zpool.ca) ($6.13)
Genoils etherum miner will give you ($4.71) of 24 hours of mining. on the same card.
Donators will get discounts on #6 when it's ready.
I have added some more quark hash:
This private miner will earn 0,00142592BTC / day @32MHASH quark. (zpool.ca) ($6.13)
Genoils etherum miner will give you ($4.71) of 24 hours of mining. on the same card.
Donators will get discounts on #6 when it's ready.
The pic was taken with -i 25 1500mhz core clock. Reduced Memclock . Gigbyte windforce 980ti g1 gaming
You can buy #5(with a boost in x11,x13.x15,quark,lyra2v2,nist5, neoscrypt) for just 0.1 btc.
is this with faster vcash and can mine decred?
Not faster. For a faster Vcash kernal. donate 0.2BTc. (15% faster than the Provos Alexis kernal) And get decred #7 included. (25-30% faster than 1.7.5)
Hi hi hehe now I can stop developing my decred mod! :-D
LOL... 
Guess you talk about SP-MOD release #4
Try with -i 29.6
i tried 1.579 but it is slower than the djrm34 mod for neoscrypt a bit slower
also more not validate nonce, djm34 version is better
also more not validate nonce, djm34 version is better
Yes but faster than release78.
i tried 1.579 but it is slower than the djrm34 mod for neoscrypt a bit slowe
also more not validate nonce, djm34 version is better
also more not validate nonce, djm34 version is better
try quark.
EDIT: so I was correct in saying that the "constant midstate" trick wasn't useful.
The constant midstate is useful. Here is the blake-512 20% faster checkin:
https://github.com/sp-hash/ccminer/commit/68484024167132cdee5e71d57566c8aa5fdd1399
I store everything in contant memory. I think it is faster to store the precalc table directly in the instructions.
but then you must make a JIT compiler. (new kernal for every midstate)
Jump to: