dudeguy
Member
Offline
Activity: 182
Merit: 10
|
|
July 11, 2013, 08:45:37 PM |
|
So what's the most optimized miner out right now?
Depends on your processors. We have ones tuned to Sandy Bridge i5 2500k, and a 620 i7 don't know what model, then one for AM FX 8320, then a generic build. I missed some for sure. Someone help me out here.
|
|
|
|
n4ru
|
|
July 11, 2013, 08:51:34 PM |
|
Any miners geared towards an i7-860?
|
|
|
|
Ethera
Member
Offline
Activity: 69
Merit: 10
|
|
July 11, 2013, 09:04:25 PM |
|
i think release5 was kinda on that.. but i would try release8 in your case, did miracles on my xeons, which is basicly the same architecture (more or less)...
|
|
|
|
n4ru
|
|
July 11, 2013, 09:07:40 PM |
|
This is my build, I used -O2 -march=native, I'm running a FX 8350, it gives me ~1250 PPS on that. I placed all the dlls in the rar. Please do your virus check and everything, I compiled that on a VM but still take precautions running a file made by another person. https://www.dropbox.com/s/6hycwi0i4dzwu1i/Primecoin.rarPossibility of a Linux binary? i think release5 was kinda on that.. but i would try release8 in your case, did miracles on my xeons, which is basicly the same architecture (more or less)...
Where is that located? I'm seeing scattered binaries all over this thread.
|
|
|
|
8bitPunk
Member
Offline
Activity: 70
Merit: 10
|
|
July 11, 2013, 09:08:13 PM |
|
do you know if it is going to cpu only or with gpu support? It is cpu focused at the moment
|
BTC 18bPunkuginRBm1Xz9mcgj8mWJnHDAW5Th | Ł LTCgXEdyBdoQ9WdF6JHi7Pa2EWtzbDjG76 | Ψ ATEBiTLkLpAYeW5hQknUfSvnb7Abbgegku
|
|
|
dudeguy
Member
Offline
Activity: 182
Merit: 10
|
|
July 11, 2013, 09:16:15 PM |
|
This is my build, I used -O2 -march=native, I'm running a FX 8350, it gives me ~1250 PPS on that. I placed all the dlls in the rar. Please do your virus check and everything, I compiled that on a VM but still take precautions running a file made by another person. https://www.dropbox.com/s/6hycwi0i4dzwu1i/Primecoin.rarWov, this thing flies. On my i3 four cores @3Ghz now have 350-400 primespersec! Original 0.1.0 client has around 50, version 5 from last nigh around 150, and now at least 350 with your version. If it is really that fast and not some error, I would mine a few blocks tonight. Will let you know. Thanks, Zalfrin! I'm finding revision 5 is working the best on my ivy bridge i3. I get over 400 with this the 5th revision. Have you tried it?
|
|
|
|
dudeguy
Member
Offline
Activity: 182
Merit: 10
|
|
July 11, 2013, 09:20:16 PM |
|
This is my build, I used -O2 -march=native, I'm running a FX 8350, it gives me ~1250 PPS on that. I placed all the dlls in the rar. Please do your virus check and everything, I compiled that on a VM but still take precautions running a file made by another person. https://www.dropbox.com/s/6hycwi0i4dzwu1i/Primecoin.rarPossibility of a Linux binary? i think release5 was kinda on that.. but i would try release8 in your case, did miracles on my xeons, which is basicly the same architecture (more or less)...
Where is that located? I'm seeing scattered binaries all over this thread. I'm going to have to second that. I'm not sure which release is suited best for i3 ivy i5 i7 and i3 i5 i7 sandy and i3 i5 i7 haswell and FX 8320 or 8350 etc. We need some organization. I've been trying to keep people up to date on reddit but to no avail. For anyone who has put up a link we need a new thread with only links to QT downloads present (and minimal comments or just OP can keep adding clients to first post of thread). Get on the donations people. MORE!
|
|
|
|
Zalfrin
|
|
July 11, 2013, 09:23:12 PM Last edit: July 12, 2013, 01:44:27 AM by Zalfrin |
|
Here's a dump of all the links, but I'm too lazy to go back through and remind myself what each one was for... the later ones have it included in their name. You can find the answers in this thread. <snip - go download the updated clients at the end of this thread, not these old ones using unofficial codebase> Note these are all builds I did. anonppcoin and another guy had a build available as well. Caveat emptor with all the builds in this thread. Thanks to the 8 people who have sent donations. Much appreciated.
|
|
|
|
dudeguy
Member
Offline
Activity: 182
Merit: 10
|
|
July 11, 2013, 09:41:56 PM |
|
Found a block. 5 XPM your way. More to come proportionate to block finding. THANK YOU!!!!!!!!!!!!!!!!!!!!!!
|
|
|
|
Ethera
Member
Offline
Activity: 69
Merit: 10
|
|
July 11, 2013, 09:52:16 PM |
|
is this another freak/nerd coin? where only a few people know how it works?
i doubt that you know how any of the coin works besides "oh it hashes babe..."... Sorry mate but its the truth. The prime numbers were covered in middle school, class 4 i guess? And if you still fail to understand (with reading how the process is going) then what the hell you are doing in sha or scrypt? or your understanding of the coin is "i got a miner that friend showed me how to run".. Seriously...
|
|
|
|
FiiNALiZE
|
|
July 11, 2013, 09:59:17 PM |
|
is this another freak/nerd coin? where only a few people know how it works?
i doubt that you know how any of the coin works besides "oh it hashes babe..."... Sorry mate but its the truth. The prime numbers were covered in middle school, class 4 i guess? And if you still fail to understand (with reading how the process is going) then what the hell you are doing in sha or scrypt? or your understanding of the coin is "i got a miner that friend showed me how to run".. Seriously... You don't need to know how it works to mine it. But since I have time, I'll show you what prime numbers are. 3 is prime because it only has two divisors (1 and itself) that can divide it into an integer. 6 is not prime because it has more than two divisors (1, 6, 2, 3). You get it now?
|
|
|
|
n4ru
|
|
July 11, 2013, 10:03:29 PM |
|
is this another freak/nerd coin? where only a few people know how it works?
i doubt that you know how any of the coin works besides "oh it hashes babe..."... Sorry mate but its the truth. The prime numbers were covered in middle school, class 4 i guess? And if you still fail to understand (with reading how the process is going) then what the hell you are doing in sha or scrypt? or your understanding of the coin is "i got a miner that friend showed me how to run".. Seriously... Which middle school did you go to that taught you Cunningham chains?
|
|
|
|
Ethera
Member
Offline
Activity: 69
Merit: 10
|
|
July 11, 2013, 10:04:19 PM |
|
is this another freak/nerd coin? where only a few people know how it works?
i doubt that you know how any of the coin works besides "oh it hashes babe..."... Sorry mate but its the truth. The prime numbers were covered in middle school, class 4 i guess? And if you still fail to understand (with reading how the process is going) then what the hell you are doing in sha or scrypt? or your understanding of the coin is "i got a miner that friend showed me how to run".. Seriously... I've written an implementation of SHA-{224,256,384,512}. wolfie that wasnt adressed to you, you are awesome part of the community and I solute you for it. it was toward another "oh i am hashing at that all others are crapcoins due me not understanding" lad there..
|
|
|
|
TheSwede75
|
|
July 11, 2013, 10:10:06 PM |
|
A list of builds and optimization as per processor family (Intel/AMD) would be useful. Right now it feels like most people are just sort of installing and guessing/trial and error style.
|
|
|
|
Zalfrin
|
|
July 11, 2013, 10:13:53 PM Last edit: July 12, 2013, 01:44:06 AM by Zalfrin |
|
Here's a dump of all the links, but I'm too lazy to go back through and remind myself what each one was for... the later ones have it included in their name. You can find the answers in this thread. <snip - go download the updated clients at the end of this thread, not these old ones using unofficial codebase> Note these are all builds I did. anonppcoin and another guy had a build available as well. Caveat emptor with all the builds in this thread. Thanks to the 8 people who have sent donations. Much appreciated. A list of builds and optimization as per processor family (Intel/AMD) would be useful. Right now it feels like most people are just sort of installing and guessing/trial and error style. Basically 3-5 were all attempts at optimizing for a sandybridge i5. 6+ has what it was optimized for in its name. 8 is probably the most generic / most likely to run for the widest audience. 5 will probably be faster if you have something in the i5 family.
|
|
|
|
fabrizziop
|
|
July 11, 2013, 10:21:18 PM Last edit: July 12, 2013, 03:14:28 AM by fabrizziop |
|
I'm making a couple of extra builds here, mainly to gain a bit of experience as I need it once in a while. I'm doing a Pentium-M, Atom, Athlon64(for the guy in the laptop) and a core 2 duo build, I'll upload them as I can. All are compiled with -O2 as if I use -O3 I get some errors with rpc stuff at the end, the file runs but it may have bad functionality. Athlon 64 https://www.dropbox.com/s/x5a7k8bave4wp3i/primecoin-athlon64.rarCore 2 Duo https://www.dropbox.com/s/x4c8q0ou4e7vil2/primecoin-core2.rarAtom https://www.dropbox.com/s/e0rsn0hzwx7h3gv/primecoin-atom.rarBarcelona (AMD K10 parts), this one is -O3, threw some errors at the end but it works fine for me. https://www.dropbox.com/s/easckkzjzlo5bt5/primecoin-barcelona.rarSandybridge (probably) https://www.dropbox.com/s/371fis1dbxfb26n/primecoin-i7-avx-i.rar
|
|
|
|
Sunny King (OP)
Legendary
Offline
Activity: 1205
Merit: 1010
|
|
July 11, 2013, 10:26:54 PM |
|
stuck at 2 blocks synced no matter how many connections i have
received block 9657908608eeddce30390a30ac72b3348fd584c1b3cb917922707f9cfc01a488 SetBestChain: new best=9657908608eeddce30390a30ac72b3348fd584c1b3cb917922707f9cfc01a488 height=1 difficulty=7 log2Work=13 log2ChainWork=13.044394 tx=2 date=2013-07-07 18:28:00 progress=0.000000 ProcessBlock: ACCEPTED received block 7edd2add999ca4a7d3c6abbe827a9aa19ea3e0e782e80c299d22599a29ec1f34 SetBestChain: new best=7edd2add999ca4a7d3c6abbe827a9aa19ea3e0e782e80c299d22599a29ec1f34 height=2 difficulty=7 log2Work=13 log2ChainWork=14.022368 tx=3 date=2013-07-07 18:29:58 progress=0.000000 ProcessBlock: ACCEPTED received block b22946daedabb93d4e9e28945278fd3d9f7ec601ce4aba1a25e6e203cf6fa41f ERROR: AcceptBlock() : incorrect proof of work ERROR: ProcessBlock() : AcceptBlock FAILED Misbehaving: 192.241.209.212:9911 (0 -> 100) DISCONNECTING
Try using -debug and -printtarget and post the ERROR messages about block #3 being rejected again This gave me a little hint. FYI sunny, the code this was built from is here, it is slightly modified from your latest commit to include some additional suggested optimizations from Mike270. https://github.com/Zalfrin/primecoinThere may be a bug in the way the "phash" variable is replacing "pblock->GetHeaderHash();". Maybe phash needs to be updated within the loop starting on line 4640 of main.cpp Please be *very careful* with protocol related code. If you break the protocol all users on a fork caused by your build will be invalid. Please uncomment the print call in TargetGetNext() and try to reproduce the issue with block#3 rejection due to wrong target. The print is only turned on when use -debug and -printtarget
|
|
|
|
annette786
|
|
July 11, 2013, 10:33:52 PM |
|
Zalfrin--
I'm using release5 with my i5 and have awesome PPS, but I've gotten stock twice lost a couple blocks. Just an FYI.
|
|
|
|
Sunny King (OP)
Legendary
Offline
Activity: 1205
Merit: 1010
|
|
July 11, 2013, 10:34:40 PM |
|
I notice some users releasing builds with changes to protocol related code without my review. This is dangerous. If you want to optimize that aggressively you should be working on an external miner instead.
Please remove any optimization code touching protocol related functions such as Target* functions, CheckProofOfWork(), ProbablePrimeChainTest() etc.
|
|
|
|
Eli0t
|
|
July 11, 2013, 10:37:24 PM |
|
I notice some users releasing builds with changes to protocol related code without my review. This is dangerous. If you want to optimize that aggressively you should be working on an external miner instead.
Please remove any optimization code touching protocol related functions such as Target* functions, CheckProofOfWork(), ProbablePrimeChainTest() etc.
this wouldnt be an issue if there was an official build with the current set of optimisations
|
LTC: LKpJf3uk7KsHU73kxq8iFJrP1AAKN7Yni7 DGC: DKXGvEbj3Rwgrm2QQbRyNPDDZDYoq4Y44d XPM: AWV5AKfLFyoBaMjg9C77rGUBhuFxz5DGGL
|
|
|
|