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

[Tuck Fheman]

hp6 is better just because I keep saying "chain spermin" repeatedly and laughing.

[1714035444]

1714035444

[1714035444]

1714035444

[FreeTrade]

and what the heck is Quarkcoin? ... doesn't exist on paper, or the internet (lolz)

Too new for Google!

[FreeTrade]

Yes, completely different.

Does Primecoin difficulty scale approximately linearly with computing power? If the processing power doubles, does the difficulty double (eventually) too?

[eCoinomist]

posted in the wrong thread, moving here:
@mikaelh, it seems to be crashing a lot with HP6:

Code:

primecoind: checkqueue.h:171: CCheckQueueControl<T>::CCheckQueueControl(CCheckQueue<T>*) [with T = CScriptCheck]: Assertion `pqueue->nTotal == pqueue->nIdle' failed.

I get a crash on 70% of machines in 5 minutes gap (16cores).

Though I did not use your Linux compliation, (used my own, since it worked better for me).

Do I need to install it your way for HP6 to work? Or perhaps Reboot? (after rebooting, seems to stopped crashing, no more error above so far)

I do notice ~15% increase in PPS from 8.3k to 9.5k pps, and chain/hr seems to average 20, just found a block,

[Effayy]

Silly question but:

What does sievepercentage refer to? If I'm going to attempt to tune it, it helps to know what exactly it's doing.

[Tuck Fheman]

Silly question but:

What does sievepercentage refer to? If I'm going to attempt to tune it, it helps to know what exactly it's doing.

It limits how many base primes are used to filter out candidate multipliers in the sieve. 8% seems optimal for testnet but I have no idea about mainnet.

[ivanlabrie]

How much does it cost to run such a vps cluster?

[notlist3d]

How much does it cost to run such a vps cluster?

Varries not all are equal.  By now most are not efficient.

[n4ru]

Anyone running many machines experienced a major slowdown in the block rate since about 14 hours ago? I was getting around 40 blocks a day, not it looks like I'm getting 20.

I'm currently averaging 14 blocks/hour after diff 9. Was at around 16 - 18/hour yesterday afternoon.

[notlist3d]

Anyone else seeing drop in PPS?

[hasle2]

Before the switch to diff 9 I was getting 400-500k pps when all the machines are added up. I started seeing the drop about 10 hours before the switch to diff 9 though. The difficulty change should not have any effect on the block rate anyway. The total average blocks per day should stay the same and I should get on average x% of the blocks where x is the % of the total mining power I have. So for my rate to drop either there was a huge jump in mining, such as many gpu miners coming online, or my setup is not working properly since 16 hours ago.

It's looking like mining is getting significantly less profitable without a gpu miner.

[Alex P]

Anyone else having issues with Crashes on Windows 7 64 bit machines and also Windows 8 64 bit machines?  It just started happenning around 6am pst today.  All versions of HPs are crashing.. up to version hp6.

Any fix??

[n4ru]

-hp6 released!

Lots of changes in this release. Here's a list of notable changes:
 * Added fast divisibility tests before doing the expensive Fermat's test
 * Lots of other small optimizations
 * Introduced a new experimental tuning parameter "sievepercentage". Default value is 10, minimum is 1, maximum is 100.
 * Added new RPC command "getchainspermin"
 * Added the following information to "getmininginfo": chainspermin, difficulty, sievepercentage, and sievesize
 * Attempt to fix the random crash while mining

My PPS on mainnet went up 23% with this release with difficulty >= 9 on mainnet. This release attempts to fix the random crash while mining. I'm not sure if it's actually fixed because I have heard both positive and negative reports. The crash is more likely with high number of cores. 16 and 32 core machines are suffering the most.

The "sievepercentage" parameter is there for people to experiment with. It limits how many base primes are used to filter out candidate multipliers in the sieve. 8% seems optimal for testnet but I have no idea about mainnet.

This build is doing horrible for me. Dropped down to 9 blocks last hour, and 5 this one (almost at the end of the hour). Was getting between 8 and 20 the last build, even after diff 9. Gonna give it one more hour of variation, then switch back. My servers restart primecoind upon crashing, so could the reduction be from multiple crashes? Any explanation?

[eCoinomist]

-hp6 released!

Lots of changes in this release. Here's a list of notable changes:
 * Added fast divisibility tests before doing the expensive Fermat's test
 * Lots of other small optimizations
 * Introduced a new experimental tuning parameter "sievepercentage". Default value is 10, minimum is 1, maximum is 100.
 * Added new RPC command "getchainspermin"
 * Added the following information to "getmininginfo": chainspermin, difficulty, sievepercentage, and sievesize
 * Attempt to fix the random crash while mining

My PPS on mainnet went up 23% with this release with difficulty >= 9 on mainnet. This release attempts to fix the random crash while mining. I'm not sure if it's actually fixed because I have heard both positive and negative reports. The crash is more likely with high number of cores. 16 and 32 core machines are suffering the most.

The "sievepercentage" parameter is there for people to experiment with. It limits how many base primes are used to filter out candidate multipliers in the sieve. 8% seems optimal for testnet but I have no idea about mainnet.

This build is doing horrible for me. Dropped down to 9 blocks last hour, and 5 this one (almost at the end of the hour). Was getting between 8 and 20 the last build, even after diff 9. Gonna give it one more hour of variation, then switch back. My servers restart primecoind upon crashing, so could the reduction be from multiple crashes? Any explanation?

doing alright for me, on one rack still getting 1-2 blocks/hr. I still have another rack on old hp5 which has 1-3 blocks/hr variable - will see after a few hours which one is better. They all crashing every often though.

[tyrion70]

-hp6 released!

Lots of changes in this release. Here's a list of notable changes:
 * Added fast divisibility tests before doing the expensive Fermat's test
 * Lots of other small optimizations
 * Introduced a new experimental tuning parameter "sievepercentage". Default value is 10, minimum is 1, maximum is 100.
 * Added new RPC command "getchainspermin"
 * Added the following information to "getmininginfo": chainspermin, difficulty, sievepercentage, and sievesize
 * Attempt to fix the random crash while mining

My PPS on mainnet went up 23% with this release with difficulty >= 9 on mainnet. This release attempts to fix the random crash while mining. I'm not sure if it's actually fixed because I have heard both positive and negative reports. The crash is more likely with high number of cores. 16 and 32 core machines are suffering the most.

The "sievepercentage" parameter is there for people to experiment with. It limits how many base primes are used to filter out candidate multipliers in the sieve. 8% seems optimal for testnet but I have no idea about mainnet.

This build is doing horrible for me. Dropped down to 9 blocks last hour, and 5 this one (almost at the end of the hour). Was getting between 8 and 20 the last build, even after diff 9. Gonna give it one more hour of variation, then switch back. My servers restart primecoind upon crashing, so could the reduction be from multiple crashes? Any explanation?

It seems that the total number of blocks found per hour changed drastically since diff 9. It now is around 3-4 blocks per minute as far as I can tell, while it was 5-6 per minute a few hours ago.. It seems difficulty is now getting where it should be.. PPS went down as well it seems; used to have 400k pps / 100 machines, now its 200k pps / 100, but still finding more than 3 blocks per 200k pps new style.. Definately getting harder though

[notlist3d]

-hp6 released!

Lots of changes in this release. Here's a list of notable changes:
 * Added fast divisibility tests before doing the expensive Fermat's test
 * Lots of other small optimizations
 * Introduced a new experimental tuning parameter "sievepercentage". Default value is 10, minimum is 1, maximum is 100.
 * Added new RPC command "getchainspermin"
 * Added the following information to "getmininginfo": chainspermin, difficulty, sievepercentage, and sievesize
 * Attempt to fix the random crash while mining

My PPS on mainnet went up 23% with this release with difficulty >= 9 on mainnet. This release attempts to fix the random crash while mining. I'm not sure if it's actually fixed because I have heard both positive and negative reports. The crash is more likely with high number of cores. 16 and 32 core machines are suffering the most.

The "sievepercentage" parameter is there for people to experiment with. It limits how many base primes are used to filter out candidate multipliers in the sieve. 8% seems optimal for testnet but I have no idea about mainnet.

This build is doing horrible for me. Dropped down to 9 blocks last hour, and 5 this one (almost at the end of the hour). Was getting between 8 and 20 the last build, even after diff 9. Gonna give it one more hour of variation, then switch back. My servers restart primecoind upon crashing, so could the reduction be from multiple crashes? Any explanation?

It seems that the total number of blocks found per hour changed drastically since diff 9. It now is around 3-4 blocks per minute as far as I can tell, while it was 5-6 per minute a few hours ago.. It seems difficulty is now getting where it should be.. PPS went down as well it seems; used to have 400k pps / 100 machines, now its 200k pps / 100, but still finding more than 3 blocks per 200k pps new style.. Definately getting harder though

I'm in PPS envy.

[CoinBuzz]

did you hard coded that path of primecoin.conf ?!

i changed the path of data folder in my run.bat so the blockchain and wallet is in some other places ( in my D drive ).

I make a primecoin.conf there and changed the sievesize to 2milion, but when i run the program, sievesize is still 1milion.

[Tuck Fheman]

It's looking like mining is getting significantly less profitable without a gpu miner.

says the guy with 400-500k pps. lulz

[notlist3d]

did you hard coded that path of primecoin.conf ?!

i changed the path of data folder in my run.bat so the blockchain and wallet is in some other places ( in my D drive ).

I make a primecoin.conf there and changed the sievesize to 2milion, but when i run the program, sievesize is still 1milion.

Same here mine says the default.

[maco]

I still don't understand how I overcame this error few days ago and it keeps coming back -- any ideas?  I tried swap commands but I guess the server I am has it locked and does not allow swapping, but it has enough memory to processor the make command: make -f makefile.unix USE_UPNP=-
I even tried make clean command and still no progress when I launch this. Any updates or ideas?

g++: internal compiler error: Killed (program cc1plus)
Please submit a full bug report,
with preprocessed source if appropriate.
See <file:///usr/share/doc/gcc-4.6/README.Bugs> for instructions.
make: *** [obj/alert.o] Error 4