[XPM] [ANN] Primecoin High Performance | HP14 released!

[MarpleTrading]

I was wondering what PPS should I get from a 12 core Xeon, using HP4?

[1686297355]

1686297355

[1686297355]

1686297355

[1686297355]

1686297355

[Vannicke]

I finally turned my intels off ... but I'm leaving one either lucky / great cpu running for kicks.

The one AMD Phenom II 1100T has netted me $300 alone, my sandy bridge i7s got me a whopping 15$ each, and a i5-3570k brought in a measly $25, not to mention this laptop, which somehow grabbed a block with its A8 ... net power usage can't have been $10 overall, since it's only been days, and the AMD is on anyhow running my LTC rig, so its hardly added to power, yet brought in loads of XPM

Yay 1$/XPM.  I made a killing this week compared to LTC.  I wonder where the bubble will pop.

[frank1892]

WOW, already getting flooded with PMs.

Take off your tinfoil hats people. I'm NOT GPU mining. I'm simply running a lot of high power private servers.

then maybe stop broadcasting it just to brag

I don't mind answering questions/getting flames/etc, but apparently everyone thinks I'm beating out people because of a GPU miner.

Would have been beneficial to you and the board if you were just more explicit.  So you have 1 vpn? 2? and on each of these what type of servers? how many? how many cores in each?  

Edit:  Also you imply VPS's are no longer profitable, when it appears you could simply duplicate what you're doing w/ vps so I'm not clear how its not equally profitable, unless you're using someone else's server in a datacenter whose resources are paid by someone else.

Amazon/DO is no longer profitable. Private companies/datacenters is something different. I am paying for the services.

May I ask how to use a lot computers calculation at one wallet? Any mining client can be use? I don't understand how can you have 600K PPS in one wallet? Please share some idea to us. Thanks.

[jubalix]

I finally turned my intels off ... but I'm leaving one either lucky / great cpu running for kicks.

The one AMD Phenom II 1100T has netted me $300 alone, my sandy bridge i7s got me a whopping 15$ each, and a i5-3570k brought in a measly $25, not to mention this laptop, which somehow grabbed a block with its A8 ... net power usage can't have been $10 overall, since it's only been days, and the AMD is on anyhow running my LTC rig, so its hardly added to power, yet brought in loads of XPM

Yay 1$/XPM.  I made a killing this week compared to LTC.  I wonder where the bubble will pop.

there going be less primes that LTC *I think* it should at least go to LTC level.....if not higher....

[bidji29]

With adjusted diff.
LTC :  28K coin/day
XPM : 18k coin/day

[flowice]

Q: Why is my performance worse on Linux?
A: This version relies on libgmp to do most of the complex calculations. Make sure your libgmp is up-to-date. Old versions do not support newer CPUs. I suggest compiling your own libgmp from latest source code for maximum performance.

What should I do?

[Krusher33]

WOW, already getting flooded with PMs.

Take off your tinfoil hats people. I'm NOT GPU mining. I'm simply running a lot of high power private servers.

then maybe stop broadcasting it just to brag

I don't mind answering questions/getting flames/etc, but apparently everyone thinks I'm beating out people because of a GPU miner.

Would have been beneficial to you and the board if you were just more explicit.  So you have 1 vpn? 2? and on each of these what type of servers? how many? how many cores in each?  

Edit:  Also you imply VPS's are no longer profitable, when it appears you could simply duplicate what you're doing w/ vps so I'm not clear how its not equally profitable, unless you're using someone else's server in a datacenter whose resources are paid by someone else.

Amazon/DO is no longer profitable. Private companies/datacenters is something different. I am paying for the services.

May I ask how to use a lot computers calculation at one wallet? Any mining client can be use? I don't understand how can you have 600K PPS in one wallet? Please share some idea to us. Thanks.

By putting the same wallet.dat on each of the clients.  Then you can just watch the transaction history from one client.

[jubalix]

i noticed in the sunny king repo he as implemented a "dynamic sieve"....is this in the hp4, is it better than hp4

[paulthetafy]

has anyone done any conclusive testing with the sievesize?  I did, and found that 2000000 was marginally better than 1000000 on the testnet, but that was using the 5-chain estimates from debug.log.  I haven't done anything on the live network and haven't heard that anyone else has either.  I think the diff is too high to do any conclusive testing now as the time between finding blocks is too high and variable.

Time for mikaelh to do some magic and put out an hp5 I think

 -PTT

[oroqen]

has anyone done any conclusive testing with the sievesize?  I did, and found that 2000000 was marginally better than 1000000 on the testnet, but that was using the 5-chain estimates from debug.log.  I haven't done anything on the live network and haven't heard that anyone else has either.  I think the diff is too high to do any conclusive testing now as the time between finding blocks is too high and variable.

Time for mikaelh to do some magic and put out an hp5 I think

 -PTT

or do the same magic with bfgminer and sort out primecoins API to work on mainnet

[eCoinomist]

has anyone done any conclusive testing with the sievesize?  I did, and found that 2000000 was marginally better than 1000000 on the testnet, but that was using the 5-chain estimates from debug.log.  I haven't done anything on the live network and haven't heard that anyone else has either.  I think the diff is too high to do any conclusive testing now as the time between finding blocks is too high and variable.

Time for mikaelh to do some magic and put out an hp5 I think

 -PTT

http://ecoinomist.com/xpm-primecoin-mining-guide-on-linux#sieve

[paulthetafy]

has anyone done any conclusive testing with the sievesize?  I did, and found that 2000000 was marginally better than 1000000 on the testnet, but that was using the 5-chain estimates from debug.log.  I haven't done anything on the live network and haven't heard that anyone else has either.  I think the diff is too high to do any conclusive testing now as the time between finding blocks is too high and variable.

Time for mikaelh to do some magic and put out an hp5 I think

 -PTT

http://ecoinomist.com/xpm-primecoin-mining-guide-on-linux#sieve

That's useful, but what about on the testnet or even looking at the 5-chains output in debug.log on the live net?  mikeal himself reported finding more blocks on the testnet with 2m

[Tamis]

http://www.coindesk.com/discount-code-and-primecoin-mining-enthusiasm-cause-cloud-server-overload/

[mhps]

http://www.coindesk.com/discount-code-and-primecoin-mining-enthusiasm-cause-cloud-server-overload/

The referral kickback probably made more money for the poster than the blocks he/she mined.

[mikaelh]

-hp5 released!

This version adds a fast 32-bit routine for inverting numbers. I did some other small performance improvements to the sieve as well. Nothing really big though. Only a 2% improvement in my benchmark.

Official backup repository is now on bitbucket:
https://bitbucket.org/mikaelh/primecoin-hp

But please download the source code from SourceForge if you can.

[blastbob]

-hp5 released!

This version adds a fast 32-bit routine for inverting numbers. I did some other small performance improvements to the sieve as well. Nothing really big though. Only a 2% improvement in my benchmark.

Official backup repository is now on bitbucket:
https://bitbucket.org/mikaelh/primecoin-hp

But please download the source code from SourceForge if you can.

Good stuff

Can you see if you can fix this also..

primecoind: checkqueue.h:167: CCheckQueueControl<T>::CCheckQueueControl(CCheckQueue<T>*) [with T = CScriptCheck]: Assertion `pqueue->nTotal == pqueue->nIdle' failed.

CCheckQueueControl(CCheckQueue<T> *pqueueIn) : pqueue(pqueueIn), fDone(false) {
        // passed queue is supposed to be unused, or NULL
        if (pqueue != NULL) {
            assert(pqueue->nTotal == pqueue->nIdle);
            assert(pqueue->nTodo == 0);
            assert(pqueue->fAllOk == true);
        }
    }

It crashes my daemons.. Should be freed or released ?

[mikaelh]

Good stuff

Can you see if you can fix this also..

primecoind: checkqueue.h:167: CCheckQueueControl<T>::CCheckQueueControl(CCheckQueue<T>*) [with T = CScriptCheck]: Assertion `pqueue->nTotal == pqueue->nIdle' failed.

CCheckQueueControl(CCheckQueue<T> *pqueueIn) : pqueue(pqueueIn), fDone(false) {
        // passed queue is supposed to be unused, or NULL
        if (pqueue != NULL) {
            assert(pqueue->nTotal == pqueue->nIdle);
            assert(pqueue->nTodo == 0);
            assert(pqueue->fAllOk == true);
        }
    }

It crashes my daemons.. Should be freed or released ?

My plan is to try out my remaining optimization ideas first. Then I'll start looking at the crashes. I haven't looked at this specific crash in detail. I have seen it happening on my boxes too and I do have a core dump of the crash. I just need some time to look at that core dump.

[blastbob]

Good stuff

Can you see if you can fix this also..

primecoind: checkqueue.h:167: CCheckQueueControl<T>::CCheckQueueControl(CCheckQueue<T>*) [with T = CScriptCheck]: Assertion `pqueue->nTotal == pqueue->nIdle' failed.

CCheckQueueControl(CCheckQueue<T> *pqueueIn) : pqueue(pqueueIn), fDone(false) {
        // passed queue is supposed to be unused, or NULL
        if (pqueue != NULL) {
            assert(pqueue->nTotal == pqueue->nIdle);
            assert(pqueue->nTodo == 0);
            assert(pqueue->fAllOk == true);
        }
    }

It crashes my daemons.. Should be freed or released ?

My plan is to try out my remaining optimization ideas first. Then I'll start looking at the crashes. I haven't looked at this specific crash in detail. I have seen it happening on my boxes too and I do have a core dump of the crash. I just need some time to look at that core dump.

Memleak?

Sent you a BTC for motivation!

[Krusher33]

I know it was mentioned somewhere in this 770 post thread but how do you update in linux again?

Or do you just delete the old and do it all over again?

[jubalix]

-hp5 released!

This version adds a fast 32-bit routine for inverting numbers. I did some other small performance improvements to the sieve as well. Nothing really big though. Only a 2% improvement in my benchmark.

Official backup repository is now on bitbucket:
https://bitbucket.org/mikaelh/primecoin-hp

But please download the source code from SourceForge if you can.

aaaand compiled and running