|
Nope, hardcopy is a "real" sASIC, I think you're thinking of xilinx easypath (which is pretty much maskROM programmed FPGA).
And I never quite said that sASIC isn't competitive, I said the specific device I used was barely competitive with 45nm FPGAs. Mainly thanks to having no dedicated adder resources and being older (but also low up-front cost) technology.
If this is based on something like a 40nm hardcopy III, performance and power figures look like they're in the right ballpark.
|
|
|
|
Yes, looks like about 10% of blocks still contain spam. 0.5.0.8 only ignores spendable outputs < mininput for sending, so if you're at the receiving end of dust it still shows up in listtransactions and friends and causes increased memory use (current bitcoind and forks keep all wallet tx in memory). If you want to try, http://pastie.org/2909400 is a band-aid to make the wallet ignore new incoming tx that only contain outputs to you < mininput. If you really want the dust later just set -mininput 0.00000001 and do a -rescan. |
|
|
|
I have some miners using linux for mining. I would need some noobs instructions for having this running. I did "Dependency Build Instructions" for debian, don't want upnp, no qt UI, I'm currently in front this page long building error, makefile.unix:131: warning: overriding commands for target `obj/nogui/scrypt.o'
makefile.unix:128: warning: ignoring old commands for target `obj/nogui/scrypt.o '
g++ -c -pthread -Wno-invalid-offsetof -Wformat -g -DNOPCH -DUSE_UPNP=0 -DUSE_SS L -fstack-protector-all -Wstack-protector -Wl,-z,relro -Wl,-z,now -D_FORTIFY_SOU RCE=2 -O2 -MMD -o obj/nogui/crypter.o crypter.cpp
In file included from crypter.cpp:9:
headers.h:36:20: warning: db_cxx.h: No such file or directory
In file included from main.h:12,
from headers.h:89,
from crypter.cpp:9:
db.h:29: error: âDbEnvâ does not name a type
db.h:42: error: ISO C++ forbids declaration of âDbâ with no type
db.h:42: error: expected â;â before â*â token
db.h:44: error: âDbTxnâ was not declared in this scope
db.h:44: error: template argument 1 is invalid
db.h:44: error: template argument 2 is invalid
db.h:157: error: ISO C++ forbids declaration of âDbcâ with no type
db.h:157: error: expected â;â before â*â token
db.h:168: error: expected â;â before âintâ
db.h:168: error: âDbcâ has not been declared
db.h:207: error: ISO C++ forbids declaration of âDbTxnâ with no type
db.h:207: error: expected â;â before â*â token
db.h:215: error: expected â;â before âpublicâ
db.h:168: error: âDB_NEXTâ was not declared in this scope
db.h: In member function âbool CDB::Read(const K&, T&)â:
db.h:59: error: âpdbâ was not declared in this scope
db.h:66: error: âDbtâ was not declared in this scope
db.h:66: error: expected â;â before âdatKeyâ
db.h:69: error: expected â;â before âdatValueâ
db.h:70: error: âdatValueâ was not declared in this scope
db.h:70: error: âDB_DBT_MALLOCâ was not declared in this scope
db.h:71: error: âpdbâ was not declared in this scope
db.h:71: error: there are no arguments to âGetTxnâ that depend on a template par ameter, so a declaration of âGetTxnâ must be available
db.h:71: note: (if you use â-fpermissiveâ, G++ will accept your code, but allowi ng the use of an undeclared name is deprecated)
db.h:71: error: âdatKeyâ was not declared in this scope
db.h: In member function âbool CDB::Write(const K&, const T&, bool)â:
db.h:89: error: âpdbâ was not declared in this scope
db.h:98: error: âDbtâ was not declared in this scope
db.h:98: error: expected â;â before âdatKeyâ
db.h:104: error: expected â;â before âdatValueâ
db.h:107: error: âpdbâ was not declared in this scope
db.h:107: error: there are no arguments to âGetTxnâ that depend on a template pa rameter, so a declaration of âGetTxnâ must be available
db.h:107: error: âdatKeyâ was not declared in this scope
db.h:107: error: âdatValueâ was not declared in this scope
db.h:107: error: âDB_NOOVERWRITEâ was not declared in this scope
db.h: In member function âbool CDB::Erase(const K&)â:
db.h:118: error: âpdbâ was not declared in this scope
db.h:127: error: âDbtâ was not declared in this scope
db.h:127: error: expected â;â before âdatKeyâ
db.h:130: error: âpdbâ was not declared in this scope
db.h:130: error: there are no arguments to âGetTxnâ that depend on a template pa rameter, so a declaration of âGetTxnâ must be available
db.h:130: error: âdatKeyâ was not declared in this scope
db.h:134: error: âDB_NOTFOUNDâ was not declared in this scope
db.h: In member function âbool CDB::Exists(const K&)â:
db.h:140: error: âpdbâ was not declared in this scope
db.h:147: error: âDbtâ was not declared in this scope
db.h:147: error: expected â;â before âdatKeyâ
db.h:150: error: âpdbâ was not declared in this scope
db.h:150: error: there are no arguments to âGetTxnâ that depend on a template pa rameter, so a declaration of âGetTxnâ must be available
db.h:150: error: âdatKeyâ was not declared in this scope
db.h: In member function âint CDB::ReadAtCursor(int*, CDataStream&, CDataStream& , unsigned int)â:
db.h:171: error: âDbtâ was not declared in this scope
db.h:171: error: expected â;â before âdatKeyâ
db.h:172: error: âDB_SETâ was not declared in this scope
db.h:172: error: âDB_SET_RANGEâ was not declared in this scope
db.h:172: error: âDB_GET_BOTHâ was not declared in this scope
db.h:172: error: âDB_GET_BOTH_RANGEâ was not declared in this scope
db.h:174: error: âdatKeyâ was not declared in this scope
db.h:177: error: expected â;â before âdatValueâ
db.h:178: error: âDB_GET_BOTHâ was not declared in this scope
db.h:178: error: âDB_GET_BOTH_RANGEâ was not declared in this scope
db.h:180: error: âdatValueâ was not declared in this scope
db.h:183: error: âdatKeyâ was not declared in this scope
db.h:183: error: âDB_DBT_MALLOCâ was not declared in this scope
db.h:184: error: âdatValueâ was not declared in this scope
db.h:185: error: request for member âgetâ in â* pcursorâ, which is of non-class type âintâ
db.h: In member function âbool CDB::TxnBegin()â:
db.h:218: error: âpdbâ was not declared in this scope
db.h:220: error: âDbTxnâ was not declared in this scope
db.h:220: error: âptxnâ was not declared in this scope
db.h:221: error: âdbenvâ was not declared in this scope
db.h:221: error: âGetTxnâ was not declared in this scope
db.h:221: error: âDB_TXN_NOSYNCâ was not declared in this scope
db.h:224: error: request for member âpush_backâ in â((CDB*)this)->CDB::vTxnâ, wh ich is of non-class type âintâ
db.h: In member function âbool CDB::TxnCommit()â:
db.h:230: error: âpdbâ was not declared in this scope
db.h:232: error: request for member âemptyâ in â((CDB*)this)->CDB::vTxnâ, which is of non-class type âintâ
db.h:234: error: request for member âbackâ in â((CDB*)this)->CDB::vTxnâ, which i s of non-class type âintâ
db.h:235: error: request for member âpop_backâ in â((CDB*)this)->CDB::vTxnâ, whi ch is of non-class type âintâ
db.h: In member function âbool CDB::TxnAbort()â:
db.h:241: error: âpdbâ was not declared in this scope
db.h:243: error: request for member âemptyâ in â((CDB*)this)->CDB::vTxnâ, which is of non-class type âintâ
db.h:245: error: request for member âbackâ in â((CDB*)this)->CDB::vTxnâ, which i s of non-class type âintâ
db.h:246: error: request for member âpop_backâ in â((CDB*)this)->CDB::vTxnâ, whi ch is of non-class type âintâ
make: *** [obj/nogui/crypter.o] Error 1
Do someone understand this gibberish ? apt-get install libdb4.8++-dev (or 4.7 or 5.1 or whatever you want) |
|
|
|
Getting transaction error: user@miner:~$ ./ggd sendfrom "" .... 10000
error: {"code":-4,"message":"Error: Transaction creation failed "} (address removed from paste) The same thing happens with sendtoaddress command my getinfo: {
"version" : 36977,
"balance" : 10095.36637813,
"blocks" : 149802,
"connections" : 8,
"proxy" : "",
"generate" : false,
"genproclimit" : -1,
"difficulty" : 34.51992560,
"hashespersec" : 0,
"Likely to be Geistnet?" : true,
"keypoololdest" : 1318209102,
"paytxfee" : 0.00000000,
"errors" : ""
}
Been mining solo for a few days, with properly forwarded ports. I tried to recompile the daemon, but the error remains. I tried to change the transaction fee to 1 GG and to 10 GG but nothing again. Any ideas? 10000 at once is too much. Assuming your inputs are block rewards (7 GG each), you're trying to create a transaction with 1429 inputs. each input is ~230 bytes -> that'd be a ~330kB tx. Stock client refuses to create transactions > 100kB. So if all you have are generations, either do 3000GG or less at a time, or if it has to be a single tx send to yourself first to consolidate your generations into a few bigger outputs. |
|
|
|
Nah, I was just riffing on the fact that ripper234 is very, very very aware of the premine  (He even started a thread 'bout it) As I said, I figured a better way to premine would be to ... you know, just mine alone with the minimum difficulty, instead of adjusting the entire blockchain math. Changing the blockchain math seems a lot more elegant than setting the genesis timestamp to 3 years ago and premining 300k blocks, requiring everyone to download and store 65MB of completely superfluous data. |
|
|
|
I still find that using the current two CPU coins, after a certain period of time, my computer with the most cores tend to behave weirdly processing very few hashes extremely quickly, have to stop and restart, whereas those with few cores are more stable. Wonder anyone having the same issues?
You know, it'd be simpler if you posted bug reports in easier to find places.  This happens with low --scantime and is inherited from upstream cpuminer. This commit should fix it: https://github.com/ArtForz/cpuminer/commit/99084f8be4de50789345c42ae1e56378496c2981 |
|
|
|
You're trying to build the gui client, for the daemon... make bitcoind and the bitcoin wx gui always needed wx 2.9 ... |
|
|
|
Ahh, you'll need to install libcurl then. After that, there will be libcurl.so at that path.
sudo apt-get install libcurl? Yup, on a recent debian/ubuntu just installing one of the libcurl4-*-dev packages should be enough. |
|
|
|
That's pretty wild, faster than the 1055T if that's with hyperthreading on. The Intel people in this thread are going to want to know what your configure options and compiler was for that.
Nope, that's with HT off. Measurably less heat & power with HT off compared to on, and less than 1% difference in performance. Latest ArtForz scrypt.c loses about 15% compared to the 2.85 high watermark with the same compiler & options on the previous version, FYI. That's some very impressive hand optimization for a particular architecture. Unfortunately even though I'm 2.8% of the tenebrix network I'm only pulling in about 1 BTC/day with the two i7s I have mining this (which still outperforms my 5830s by a factor of 4 in terms of power/$) so it'll be a while till I feel like buying an x6 phenom to try the same compiler/options/minerd version bake-off with. Sad to hear that, as the current HEAD also improves speeds for K10 even when compiled with older gcc versions to near gcc-4.6.1 levels Guess this one will have to stay in non-official state for now (looks like it also produces rather crap asm when compiled for 32 bit targets...). |
|
|
|
Not sure what's going on there (aka "can't reproduce issue").
My only guess is it's possibly related to removing the sha256 algos, but... that was even before I started doing the compilers job for scrypt.
Not sure what to do other than general hints along the lines of "start with a clean tree, CFLAGS="-whatever" ./configure; make"
Hrrrm... I guess you could revert the sha256 removal or drop the new scrypt.c into Tenebrix-miner and see if that also causes the same issues.
There is no scrypt.c in your latest released source of 1.0.2... I'm looking at the tar.gz and it's missing it. What release? The releases are done by Lolcust, I only keep a git at https://github.com/ArtForz/cpuminer |
|
|
|
Shrug, don't care much. But this is interesting: So Mr. "Professional Developer" has a hobby of stalking perceived "enemies" on the internet and boasting about it... *facepalm* |
|
|
|
|
Well, technically 7 million is less than one hundred thousand thousand.
But yeah, I agree. the 7.7mil superfund should be laundry-only to prevent abuse.
|
|
|
|
I am almost certain they are three separate people, especially Artforz.
Yes it's easy to make out the man dressed in animal costumes. When you have no product to show and no arguments, try some good old mudslinging. That's totally the behavior you'd expect from a "professional developer" with "decades of experience creating commercial software", not some 22yo aspie with a superiority complex living in his moms basement, right? See, easy! |
|
|
|
|
I've run 3*5970 on this exact board, so > 4 GPUs does work.
Yep, PSU is a possiblility, there's 2 current Rosewill 1kW units:
RBR-1000M = Andyson OEM. Had half a dozen 1kW PSUs based on the same Andyson platform crap out after ~3 months while only running dual 5970s.
LIGHTNING-1000 = SuperFlower OEM. Seems pretty good, so far had zero of those die, and thats with 2*5970+2*5770 or 3*5970 on em.
Anything older... not sure, probably not too good.
|
|
|
|
Sounds like a damn great idea.
I agree the KISS version should be relatively easy to implement, but I suspect it'll become dog-slow if the laundry ever becomes popular.
Basically, if a tx input is spending a output belonging to one of the 2 laundry addresses, sum up *all* unspent outputs of those 2 and consider the tx invalid if spending that output would bring the total below the threshold.
The problem is when the deposit buffer fills up with 10000s of tx... that scan-and-sum needs caching.
Should be easy to keep track of a running total, right? When it initially load the blocks, it calculates the coins in those addresses the same way it calculates your unspent coins in your addresses. So at each block, you always know how many coins those 2 addresses contains. It will be similar to how the client won't let you spend coins you don't own. Should be simple enough. And it should be simple enough to test it also. Once the code is out there, Lolcust can create a transaction that tries to spend coins that violates this rule and people can test to make sure that the transaction is indeed rejected by the code. True. Just wondering how "fun" handling this on reorgs will be. Thats why I went with the "start with the stupidest possible implementation, optimize later" assumption |
|
|
|
Basically people are wary about the 7.7 million premined coins. It takes 3 years for the network the mine another 7 million coins. So for the foreseeable future, Lolcust has absolute control of the market. And if he decides to unload part of his laundry fund in the future, no one can stop him. Or if some malicious user manages to hack into his computer to steal his coins, they can easily crash the market. And that's troublesome.
Here's a suggestion I made to Lolcust on IRC. Basically have the Tenebrix code enforce that the laundry fund can never be stolen by Lolcust or anyone else for that matter. The laundry can be implemented using 2 buffers: clean buffer and deposit buffer. These buffers are just 2 tenebrix addresses that contain coins. The clean buffer will be where clean coins come out of. The deposit buffer will be where people send their coins that they want cleaned. After a random amount of time, the same amount (minus fees) will be sent to them from the clean buffer. Initially, the clean buffer will contain the 7 million coins and the deposit buffer will have no coins. After 7 million coins are laundered, the clean buffer will run out, and the 2 buffers will swap roles.
Given this implementation, we know that if the laundry is running correctly, the sum of the 2 buffers will always be greater than 7 million coins. If at any point in time there are less than 7 million coins in the 2 buffers, then either there's a mistake in the laundry code or someone is deliberately stealing from the laundry. My proposal is that we change the code to prevent this from ever happening. All you need to do is add an extra check in the transaction checking code. If the transaction involves taking money out of either of these buffer, just make sure that the 2 buffer still sums up to at least 7 million. If it doesn't then the transaction will be considered illegal and will not be added to a block and confirmed. And if a rogue miner tries to add an illegal transaction to a block it found, that block will be rejected by other miners. With this code, people can be sure that this 7 million premined coins cannot be used for anything other than the laundry and will be forever stuck in those two addresses. And in the future, if Lolcust turns rogue and releases code without this check, people will see that and refuse to update their client to that one.
I offered to help Lolcust code this up. What do people think?
Sounds like a damn great idea. I agree the KISS version should be relatively easy to implement, but I suspect it'll become dog-slow if the laundry ever becomes popular. Basically, if a tx input is spending a output belonging to one of the 2 laundry addresses, sum up *all* unspent outputs of those 2 and consider the tx invalid if spending that output would bring the total below the threshold. The problem is when the deposit buffer fills up with 10000s of tx... that scan-and-sum needs caching. |
|
|
|
Just pushed another small tweak, gets another 3% or so on K10s.
Hey, For some reason after compiling the new code, the program no longer takes command line arguments? Not sure what's happening. It just returns the -h line no matter what I input. Also, I've been running the last version of your code at ~38kh/s and haven't gotten any blocks in about 6 hours. But maybe I'm just unlucky. Not sure what's going on there (aka "can't reproduce issue"). My only guess is it's possibly related to removing the sha256 algos, but... that was even before I started doing the compilers job for scrypt. Not sure what to do other than general hints along the lines of "start with a clean tree, CFLAGS="-whatever" ./configure; make" Hrrrm... I guess you could revert the sha256 removal or drop the new scrypt.c into Tenebrix-miner and see if that also causes the same issues. |
|
|
|
|
Just pushed another small tweak, gets another 3% or so on K10s.
|
|
|
|
|
Show Posts
