[XPM] [ANN] Primecoin High Performance | HP14 released! - page 114.

kcanup

legendary

Activity: 1484

Merit: 1000

{
"blocks" : 48294,
"currentblocksize" : 1000,
"currentblocktx" : 0,
"errors" : "",
"generate" : true,
"genproclimit" : -1,
"primespersec" : 3195,
"pooledtx" : 0,
"testnet" : false
}

Thats getmininginfo details.. so this means my qt client is mining primecoin? and if its mining, how long time it will take to find a block on current pps? Thanks. Wink

Krusher33

newbie

Activity: 37

Merit: 0

Prior to this release I was getting about 1400 PPS in the Windows client and about 1800-2000 PPS in my ubuntu VM.

I downloaded this Windows one and get about 5k PPS on 7 cores on my FX-8350 @ 4.7ghz.

I tried the the linux one and I'm only getting ~996 PPS?

I like using the linux one because it always seems to be 10c cooler and prior to this I was getting higher PPS in linux.

Steps I did:

Code:

apt-get install make
apt-get install git
git clone https://github.com/mikaelh2/primecoin
apt-get install libgmp-dev
sudo apt-get update
sudo apt-get install python-software-properties screen python-rrdtool python-pygame python-scipy python-twisted python-twisted-web python-imaging build-essential libglib2.0-dev libglibmm-2.4-dev python-dev autoconf automake ncurses-dev sysstat gcc-mingw32 libmysql++-dev cloog-ppl build-essential automake gcc libevent-dev libmemcached-dev libcurl4-openssl-dev zlib1g-dev libjansson-dev curl memcached libtool unzip freeglut3-dev libxi-dev libxmu-dev build-essential freeglut3-dev libxi-dev libxmu-dev mpich2 libdb-dev libminiupnpc-dev libboost-all-dev libdb++-dev git uthash* libgmp3-dev

cd ~/primecoin/src
make -f makefile.unix
mv primecoind /usr/local/bin

primecoind setgenerate true 7

I'm a linux noob so go easy on me.

Vorksholk

legendary

Activity: 1713

Merit: 1029

I'm a bit confused.

The primes we are finding are huge, what purpose does the up-to-2,000,000 or so prime sieve serve?

dudeguy

member

Activity: 182

Merit: 10

Quote from: mika1977 on July 15, 2013, 06:09:12 AM

Around 100 hours without a single block this with 20k+ PPS. I try every build. Now with latest - hp3
Please somebody paste your config file, may be there is my mistake.

Delete your blockchain and redownload after every new compile or downloaded qt. I hadn't had a block in over 48 hours at 4000 PPS. Deleted block chain on all my computers. Voila. 2 blocks last night. I may as well have mined salad because I was on a fork for 2 days.

altsay

sr. member

Activity: 359

Merit: 250

Quote from: 8bitPunk on July 15, 2013, 06:06:51 AM

Quote from: dego on July 15, 2013, 05:58:25 AM

compiled and running on Ubuntu 13.04 64bit on a Phenom 1090 hexacore, getting around 4500 - 5000 pps (as before).

What should thelatest changements result in?

I wouldn't expect any increase in PPS resulting from this change (remember PPS is not a good performance metric)... The efficiency is in how quickly the sieve is setup, ready to weave then find the prime chains. Within the gensieveroundlimitms parameter (default 1000 milliseconds) primecoind should spend less time setting up the sieve and more time finding the primes before it moves onto the next round.

So you'd suggest -sievesize=1000000 ?

fabrizziop

hero member

Activity: 506

Merit: 500

Hey, I built the last client, I'd like to know if I should be seeing a PPS increase with it, or is it just other optimization that make the client more likely to find a block?.

mika1977

full member

Activity: 364

Merit: 100

how to understand this, is there something wrong about it?

Quote

14:35:12

getaddednodeinfo [node]
Returns information about the given added node, or all added nodes
(note that onetry addnodes are not listed here)
If dns is false, only a list of added nodes will be provided,
otherwise connected information will also be available. (code -1)

de_xt

hero member

Activity: 602

Merit: 535

Account hacked from Oct 11th to Nov 1st 2017

A pool is a must at this difficulty level, IMHO, provided a working standalone miner is released. People would get less frustrated because of their bad luck at solo mining.

mikaelh

sr. member

Activity: 301

Merit: 250

Quote from: bbxx on July 15, 2013, 06:44:36 AM

There is something strange with this coin/daemon
It is locking sometimes mining nothing and using whole cpu power
To repair it kill daemon do -rescan and start again
You can easily spot this if you have immature coins locked at specific number of confirmations, but daemon block number is increasing

I wasted many cpu power because of that

I ma using hp2 build, ubuntu

Sounds a bit like your client got stuck on the wrong fork. Sunny King's original client also suffers from that issue.

bbxx

hero member

Activity: 728

Merit: 500

cryptoshark

There is something strange with this coin/daemon
It is locking sometimes mining nothing and using whole cpu power
To repair it kill daemon do -rescan and start again
You can easily spot this if you have immature coins locked at specific number of confirmations, but daemon block number is increasing

I wasted many cpu power because of that

I ma using hp2 build, ubuntu

mika1977

full member

Activity: 364

Merit: 100

this is my conf. file
do i have to add sievesize=1000000

Quote

rpcuser=1234
rpcpassword=123
rpcport=8001
gen=1
server=1
daemon=1
rpcallowip=127.0.0.1
addnode=137.116.230.21
addnode=54.225.81.214
addnode=54.225.116.178
addnode=54.225.123.48
addnode=107.20.202.112
addnode=54.225.123.87
addnode=54.225.81.214
addnode=54.225.116.178
addnode=54.225.123.48
addnode=54.225.123.87
addnode=107.20.202.112
addnode=54.224.142.224
addnode=54.224.28.160
addnode=184.72.193.229
addnode=54.225.19.196
addnode=50.16.142.10
addnode=23.20.77.101
addnode=54.224.95.193
addnode=54.234.90.173
addnode=23.21.2.143
addnode=23.20.0.137
addnode=54.227.57.231
addnode=54.227.81.48
addnode=54.224.112.133
addnode=54.224.194.112
addnode=50.16.112.122
addnode=54.214.144.235
addnode=54.244.76.225
addnode=54.245.7.222
addnode=54.244.204.96
addnode=54.214.187.152
addnode=50.112.197.152
addnode=54.214.168.133
addnode=54.214.100.229
addnode=54.244.199.134
addnode=54.245.138.68
addnode=54.214.59.206
addnode=54.244.208.77
addnode=54.214.103.110
addnode=94.23.215.174
addnode=87.98.146.72
addnode=87.98.146.72
addnode=94.23.215.174
gen=1

8bitPunk

member

Activity: 70

Merit: 10

Quote from: mika1977 on July 15, 2013, 06:09:12 AM

Around 100 hours without a single block this with 20k+ PPS. I try every build. Now with latest - hp3
Please somebody paste your config file, may be there is my mistake.

The config files are pretty hard to mess up, and even without one you can do: "setgenerate true"
and it will be mining with all the defaults.

What we are all seeing is the amount of computation required to generate a prime chain now that the difficulty is 8.8.

itod

legendary

Activity: 1974

Merit: 1077

^ Will code for Bitcoins

Quote from: mika1977 on July 15, 2013, 06:09:12 AM

Around 100 hours without a single block this with 20k+ PPS. I try every build. Now with latest - hp3
Please somebody paste your config file, may be there is my mistake.

On Linux:

Code:

gen=1
rpcuser=primecoinrpcuser
rpcpassword=primecoinrpcpassword

On Win: no config file, starting the mining manually with setgenerate. If you have config file just delete it, maybe there really is something wrong in it, but i doubt it, probably just bad luck.

dego

sr. member

Activity: 399

Merit: 250

Quote

rpcuser=username
rpcpassword=password
sievesize=1000000
gen=1

As I understand it right, adding sievesize=1000000 shouldn't be necessary as 1M is already the default sievesize...

I had 5 blocks yesterday during ~15hrs mining at 4k - 5k pps. Today no blocks have been yet found during more than 3hrs @ 4.5k pps.

mika1977

full member

Activity: 364

Merit: 100

Around 100 hours without a single block this with 20k+ PPS. I try every build. Now with latest - hp3
Please somebody paste your config file, may be there is my mistake.

dego

sr. member

Activity: 399

Merit: 250

ok. So it will be very hard to see if there are perfomance inprovements on the real network...

8bitPunk

member

Activity: 70

Merit: 10

Quote from: dego on July 15, 2013, 05:58:25 AM

compiled and running on Ubuntu 13.04 64bit on a Phenom 1090 hexacore, getting around 4500 - 5000 pps (as before).

What should thelatest changements result in?

I wouldn't expect any increase in PPS resulting from this change (remember PPS is not a good performance metric)... The efficiency is in how quickly the sieve is setup, ready to weave then find the prime chains. Within the gensieveroundlimitms parameter (default 1000 milliseconds) primecoind should spend less time setting up the sieve and more time finding the primes before it moves onto the next round.

dego

sr. member

Activity: 399

Merit: 250

compiled and running on Ubuntu 13.04 64bit on a Phenom 1090 hexacore, getting around 4500 - 5000 pps (as before).

What should thelatest changements result in?

8bitPunk

member

Activity: 70

Merit: 10

mikaelh is awake and coding Grin

Latest commit: Precompute two's inverses for all base primes. Idea taken from luke-jr

If you're on linux and compiling from the repo:

Quote from: rethaw on July 14, 2013, 11:28:55 PM

Remember folks, if you're doing this, keep your primecoind updated:

Code:

cd ~/primecoin
git pull
cd src
make -f makefile.unix USE_UPNP=-
primecoind stop
sudo mv /usr/local/bin/primecoind /usr/local/bin/primecoind.old
sudo mv primecoind /usr/local/bin/primecoind
primecoind --daemon

mhps

hero member

Activity: 516

Merit: 500

CAT.EX Exchange

Quote from: 8bitPunk on July 15, 2013, 03:42:05 AM

Quote from: mhps on July 15, 2013, 02:40:36 AM

Quote from: mikaelh on July 14, 2013, 01:47:21 PM

New version is out. It mainly adds a tuning parameter called -sievesize. Use it like this: primecoind -sievesize=2000000

@mikaelh to make sure ... the change of sievesize is not what Sunny was talking about in https://bitcointalksearch.org/topic/m.2716964 , right?

I have reduced sievesize from 2.5M to 50000 in several steps and the PSS indeed rises (not linearly). But I hope it's not just an illusion.

Sieve size isn't directly what Sunny was talking about, but it is related. If you lower the amount of work you perform or the shorten the time you spend working on it before giving up, then you will throw away a lot of potential solutions unnecessarily.

Lowering sieve size to 50,000 will raise PPS but it is an illusion.

Thanks for the explanation and for making the HP client. It's hard to determine where the illusion starts.. all I see is more PPS with lower sievesize.

Topic: [XPM] [ANN] Primecoin High Performance | HP14 released! - page 114. (Read 397616 times)