[XPM] [ANN] Primecoin Release - First Scientific Computing Cryptocurrency

[LazyOtto]

Anytime I use more than one thread my pps goes down.

Yep, I was absolutely seeing this. (On OS X.)

Any time using more than one thread I was seeing a pps of 0-3.
More often 0 or 1 than 2 or 3.

Running a single thread I've seen upwards of 35.

So, I'm now running a separate instance of primecoind for each core and I'm seeing 'reasonable' pps reported from each instance.

2013-07-10 16:34:27 primemeter    129136 prime/h    966899 test/h
2013-07-10 16:35:31 primemeter    124329 prime/h    970444 test/h
2013-07-10 16:36:36 primemeter    135273 prime/h   1089035 test/h
2013-07-10 16:37:40 primemeter    108500 prime/h    848492 test/h
2013-07-10 16:38:44 primemeter    114016 prime/h    900447 test/h
2013-07-10 16:39:49 primemeter    119136 prime/h    932809 test/h
2013-07-10 16:41:49 primemeter    123106 prime/h   1013342 test/h
2013-07-10 16:42:50 primemeter    128928 prime/h   1064533 test/h

-- edit

Relevant to several of the posts on the last two or three pages:

btw, I'm merely reporting this because it is trivial and it irritated the hell out of me last night. (In fact, last night I reached the conclusion that this is a pile of shit and not worth my incremental electricity cost.)

However, if I had done something massively non-trivial, like reprogrammed and reflashed some of the FPGA's sitting around here and started getting a block a minute, no, I would probably not just be tossing that out to the community as a freebie.

[1714222938]

1714222938

[1714222938]

1714222938

[1714222938]

1714222938

[1714222938]

1714222938

[Buffer Overflow]

People are getting upset because some people not sharing their compiled code are damaging their own reputation and the reputation of the coin. Also, it's just flat out unethical to steal blocks. 'Of course they are! This is the internetz where anything goes!' is the typical argument, however I'd like to think the crypto-community and particularly XPM miners are for the most part much better than the worst posters on the /b forums. Right now there are a lot of people proving me wrong. All you have to do is look at the statistics.

Everyone stop mining!

Dudeguy has claimed all future blocks for himself. Anyone discovering 'his' blocks is now classed as STEALING, and will feel the full force of the law!

[Zalfrin]

Hrm, well my self-compiled primecoind doesn't seem to be working quite right... It's using the CPU cycles, but doesn't seem to have noticed there are new blocks. Current block is in the 8400s. 

C:\bitcoin-0.8.3-EasyWinBuilder-v0.1\primecoin\src>primecoind getmininginfo
{
    "blocks" : 8399,
    "currentblocksize" : 1000,
    "currentblocktx" : 0,
    "errors" : "",
    "generate" : true,
    "genproclimit" : -1,
    "primespersec" : 175,
    "pooledtx" : 0,
    "testnet" : false

[bidji29]

with variance, it's hard to know if a change in the code is good or bad.
And PPS mean nothing apparently.
It's hard to optimise.


Am I the only one who search in the code instead of complaining?

I think it's in prime.h.
I think the code search the blockheader 1 time for each type of chain (cunigham1, 2 and twin).
We need to find a means to do the 3 at the same time. What do you think?

Some people have already did it, am I on a good path?

[xeroc]

theres an update on the repo

read: https://bitcointalk.org/index.php?topic=253662.0

[shakezula]

Blocks are now in the 17.4 range, just found another on my i7 after adding 4way=1 to my .conf, don't know if it helped or not, could have just been that I restarted the wallet, could be lucky.

[drawingthesun]

Blocks are now in the 17.4 range, just found another on my i7 after adding 4way=1 to my .conf, don't know if it helped or not, could have just been that I restarted the wallet, could be lucky.

what does 4 way mean?

[shakezula]

Blocks are now in the 17.4 range, just found another on my i7 after adding 4way=1 to my .conf, don't know if it helped or not, could have just been that I restarted the wallet, could be lucky.

what does 4 way mean?

From the bitcoin.conf readme:

# Use SSE instructions to try to generate bitcoins faster.
4way=1

Now, not sure if it makes a difference or not, because I'm not watching primespersec as that number seems misleading. Could be luck, could help, I was going to try and recompile with -march=corei7-avx, but then though blah and went for it.

[Zalfrin]

Blocks are now in the 17.4 range, just found another on my i7 after adding 4way=1 to my .conf, don't know if it helped or not, could have just been that I restarted the wallet, could be lucky.

what does 4 way mean?

From the bitcoin.conf readme:

# Use SSE instructions to try to generate bitcoins faster.
4way=1

Now, not sure if it makes a difference or not, because I'm not watching primespersec as that number seems misleading. Could be luck, could help, I was going to try and recompile with -march=corei7-avx, but then though blah and went for it.

anyone know if -march=corei7-avx is supported on core i5's, or what the recommended architecture setting would be for them? I'm using -march=native, I suppose that's probably fine.

[Zalfrin]

Why  +USE_UPNP:=1 ?  I thought setting this flag to 0 made it embed PNP (hence faster and more distributable to other machines).

Hrm, well my self-compiled primecoind doesn't seem to be working quite right... It's using the CPU cycles, but doesn't seem to have noticed there are new blocks. Current block is in the 8400s. 

C:\bitcoin-0.8.3-EasyWinBuilder-v0.1\primecoin\src>primecoind getmininginfo
{
    "blocks" : 8399,
    "currentblocksize" : 1000,
    "currentblocktx" : 0,
    "errors" : "",
    "generate" : true,
    "genproclimit" : -1,
    "primespersec" : 175,
    "pooledtx" : 0,
    "testnet" : false

Changing USE_UPNP back to ?=- (whatever that means) seems to have fixed this issue.

[CryptoBeggar]

For Linux users:


Is there any way to get the QT (and miner) running without running X? I would like to restart this thing through ssh.

[shakezula]

For Linux users:


Is there any way to get the QT (and miner) running without running X? I would like to restart this thing through ssh.

Just put gen=1 in your .conf file and run the ./primecoind instead of ./primecoin-qt.

[nmersulypnem]

Why  +USE_UPNP:=1 ?  I thought setting this flag to 0 made it embed PNP (hence faster and more distributable to other machines).

Hrm, well my self-compiled primecoind doesn't seem to be working quite right... It's using the CPU cycles, but doesn't seem to have noticed there are new blocks. Current block is in the 8400s. 

C:\bitcoin-0.8.3-EasyWinBuilder-v0.1\primecoin\src>primecoind getmininginfo
{
    "blocks" : 8399,
    "currentblocksize" : 1000,
    "currentblocktx" : 0,
    "errors" : "",
    "generate" : true,
    "genproclimit" : -1,
    "primespersec" : 175,
    "pooledtx" : 0,
    "testnet" : false

Changing USE_UPNP back to ?=- (whatever that means) seems to have fixed this issue.

Did anyone else get the "QtCore4.dll is missing" when executing the build qt exe?  It is also suspiciously small.    I found a reference to a similar issue from a bitcoin dev post that the qt linker was confusing /bin/ and /lib/, but those lines are no longer in the .pro file.... so I'm lost.  I really hate env issues.  I JUST WANT TO MODIFY THE SOURCE.

[captainfuture]

is there a block explorer ?

[Scott J]

is there a block explorer ?

Possibly here: http://www2.crypto-expert.com/primecoin/index.php

[LazyOtto]

theres an update on the repo

read: https://bitcointalk.org/index.php?topic=253662.0

For OS X miners:

Still seems I still have to run discrete instances per core, however, after grabbing
the update and *also* changing the makefile.osx to:
#CFLAGS = -mmacosx-version-min=10.5 -arch i386 -O3
CFLAGS = -mmacosx-version-min=10.7 -O3 /
         -arch x86_64 -funroll-loops -mfpmath=sse -fomit-frame-pointer

I'm getting substantially higher 'benchmark' numbers:
2013-07-10 17:31:58 primemeter    118780 prime/h    834911 test/h
2013-07-10 17:33:00 primemeter    145659 prime/h    963936 test/h
2013-07-10 17:34:02 primemeter    134060 prime/h    946472 test/h
2013-07-10 17:35:05 primemeter    233814 prime/h   1670377 test/h
2013-07-10 17:37:06 primemeter    256114 prime/h   1989032 test/h
2013-07-10 17:38:08 primemeter    215987 prime/h   1574243 test/h

That's from a single primecoind instance log.

YMMV.

-- edit

Of course, you can't build for / run 64bit (x86_64) unless your hardware and OS version supports 64bit.

[LazyOtto]

For Linux users:


Is there any way to get the QT (and miner) running without running X? I would like to restart this thing through ssh.

No.

Then don't use QT.

Use primecoind directly.

[keeron]

This might have already been answered (I thought I saw a reference here or on reddit)

My transaction on DigitialOcean VPS is currently in the category of "orphan". The other transactions (other machines) are all in the category of "immature". Have I lost those coins?

I tried dumping the privkey and importing it back to "home" wallet, but that didn't seem to work.

[LazyOtto]

Bummer.

Over ten minutes now running with the new, much faster, primecoind and I *still* haven't found a block.


 

[LazyOtto]

This might have already been answered (I thought I saw a reference here or on reddit)

My transaction on DigitialOcean VPS is currently in the category of "orphan". The other transactions (other machines) are all in the category of "immature". Have I lost those coins?

I tried dumping the privkey and importing it back to "home" wallet, but that didn't seem to work.

Google is your friend.

Research with keywords: bitcoin orphan