 Show Posts
|
|
Pages: « 1 2 [3] 4 5 6 »
|
Already filed an issue for that: https://github.com/fancyIX/sgminer-phi2-branch/issues/150Personally I don't like that coin just because their logo looks so unprofessional. Also, I cannot see a direct benefit of applying cross lane instructions to do the shuffling. But will spend more time on it. |
|
|
|
wow, lots of shuffling in that algo. That's what I like most -- apply GCN cross lane instructions to speed it up. |
|
|
|
Miner 5.6.1.3.b5h driver Adrenalin 18.6.1 RX 470 4GB 1100/1900
x22i 3.6 Mhs
ethash 26.25
lyra2Zz 1.4 Mhs
allium 3.0 Mhs (no change)
phi2 2.6 Mhs (no change)
lyra2Z 1.4 Mhs (no change)
lyra2rev2 22.7 Mhs
P.S. On the ethash algorithm on the same cards (in the same settings) Claymore's Dual Miner 12 gives 28.2 Mhs. Good job!
Yeah... Still slower than commercial miners. Since the main purpose of this project is open source, hope people can still get benefit from it. |
|
|
|
Recently, there have been a lot of rumors that Ethereum will move to a new hashing algorithm. Maybe you decide to make a miner under ProgPOW
ProgPow seems a good algo to apply gcn dpp instructions. Thanks for the info. Will work on it. Do you have any plans on adding x21s and x16rt? Took a quick look on those two. x21s has one step of lyra2v2, which kind of suitable for applying GCN optimization. I don't like to continue working on lyra2v2 related algorithms because there is a bin file originally from lyclminer. That bin file is fully optimized. Some closed source miners simply copy that file to use it as their own. kind of like IP theft but I am not sure how illegal that is. As to x16rt, seems no lyra2 related steps. Not sure if other steps can be optimized by GCN cross lane instructions. |
|
|
|
Recently, there have been a lot of rumors that Ethereum will move to a new hashing algorithm. Maybe you decide to make a miner under ProgPOW
ProgPow seems a good algo to apply gcn dpp instructions. Thanks for the info. Will work on it. |
|
|
|
Hey Fancy -- I've got a semi problem here... Trying to run the new Lyra2Zz algo...
If I run the bat file with your info in it, I get roughly 1.5 mhs per card... If I run the program through my 3rd party program, I end up hashing at about 1/3 of the speed.
The 3rd party app, basically makes bat files on the go as that algo is selected... Here is the info I give the program
-k lyra2zz -o stratum+tcp://[Server]:[Port] -u [User] -p [Password] -w 256 -I 19
Did you copy "bin-kernel" folder to the same folder where "sgminer" sits? And if you use -I 23 it may run faster. Winner Winner! Thank you sir! Let me know if it runs stable or not. It seems running on windows is less stable than on linux. But if it works for you then no worries. |
|
|
|
Hey Fancy -- I've got a semi problem here... Trying to run the new Lyra2Zz algo...
If I run the bat file with your info in it, I get roughly 1.5 mhs per card... If I run the program through my 3rd party program, I end up hashing at about 1/3 of the speed.
The 3rd party app, basically makes bat files on the go as that algo is selected... Here is the info I give the program
-k lyra2zz -o stratum+tcp://[Server]:[Port] -u [User] -p [Password] -w 256 -I 19
Did you copy "bin-kernel" folder to the same folder where "sgminer" sits? And if you use -I 23 it may run faster. |
|
|
|
So, Mister FancyIX, I saw you on discord in the MTP / Zcoin chatrooms. Any chance you'd be working on an implementation for AMD hardware? Andrucrypt seems to have hit a wall, Teamred seem too busy with real life stuff... and Zawawa has disappeared, allegedly to deal with greater things. That doesn't leave many AMD devs around to push the boundaries. How are you getting on?  Seems like djm is working on sgminer for this MTP thing. I plan to do some improvement based on his work. |
|
|
|
Hi, can you PM me your contact detail about implementing a new algo?
I tried to pm you but not allowed...
You can ping me in discord. ID: fancyIX |
|
|
|
I've been trying this in hiveos for x22i but there's no x22i in the drop down for sgminer your sgminer fork is listed tho so instead I used the custom one and sgminer will only use 4 cards I tried different settings with no luck so I ssh into my rig and ran sgminer and all 5 cards worked I dont understand why it wont use all cards when starting from the web dashboard
Yeah, hiveos has some problem in the drop down menu. They seems are fixing it. |
|
|
|
Hey FancyIX Thanks for this "attempt"... Phi2 isn't really living up to expectations though. RX570's just scratch 3MH/s at 1300/1900 (XMR mem straps) and Vega56@64 only get about 6.3MH/s at 1408/1100. To give you a comparison, a regular GTX 1080 will do 8 MH/s, slightly overclocked, at 85% PL... The Vegas should be equaling that at least. Currently they are hashing 22% slower and draw 35% more! Keep chugging! Yeah... still long way to go before profiting, though according to current market price most miners won't profit at all... |
|
|
|
Supports Windows and Linux. Supports windows driver and Linux AMDGPU Pro driver. Supports most cards. Comes with pre-built binary kernels to speed up Ellesmere and Vega cards. https://github.com/fancyIX/sgminer-phi2-branch/releasesEstimated speed (beta5fp4): phi2, rx 580: 3750khs; vega 56: 6750khs lyra2Z, rx 580: 1950khs; vega 56: 3600khs allium, rx 580: 4500khs, vega 56: 8200khs x22i, rx 580: 5100khs, vega 56: 9000khs I know that may be below the speed of some closed-source miners. Hope one day it can catch up. Just made a custom package for hiveos: https://github.com/fancyIX/sgminer-fancyIX-HiveOS-custom/releases |
|
|
|
you could try other letters, such as g, h, i, or maybe even j... or any other letter , or maybe even a number, but it's really hard for me to take this seriously when just keep adding ++++. will your next version be: Version ++++++++++++++++++++  I really appreciated that you are improving on the miner, but your version naming schema leaves much to be desired. Thanks. Let me modify the release page, changing 5f++++ to 5fp4 |
|
|
|
|
