did we all see the PPS drop
I went from 14k to 8k....or is that hp6 v hp5 effect?
Topic: [XPM] [ANN] Primecoin High Performance | HP14 released! - page 81. (Read 397645 times)
HP6 dropped me to 800 pps instead of 3k pps normally.
difficulty is rising fast ! gpu miner out ?
It is very much like comparing apples and oranges. For Primecoin, the integer part of difficulty determines the required probable prime chain length. E.g. if difficulty is 9.07233632, a probable prime chain of length 9 is required. The fractional part is used to determine how many length-9 chains are accepted. It scales logarithmically so that going from 9.8 to 9.9 is a lot harder than going from 9.1 to 9.2. The details for the fractional part are in Sunny King's paper.
The current algorithm also does more work for longer target chain lengths. That's why PPS numbers are also dropping.
The current algorithm also does more work for longer target chain lengths. That's why PPS numbers are also dropping.
Thanks Mikael, I read that it takes a 30 fold increase in processing power to increase difficulty by 1 . . . if so difficult to imagine difficultly ever climbing beyond 12 or 13.
999/(13*13) = 5.9 per block
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 for instructions.
make: *** [obj/alert.o] Error 4
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
make: *** [obj/alert.o] Error 4
RAM.
Does reducing the # of threads increase chainspermin for anyone else?
On my i7 I started using 6 threads and saw 8-12 CPMs.
When I switched to -1 it dropped to 3-6 CPMs.
I changed back to 6 and it increased again.
On my i7 I started using 6 threads and saw 8-12 CPMs.
When I switched to -1 it dropped to 3-6 CPMs.
I changed back to 6 and it increased again.
I wasn't able to replicate this. Maybe this has something to do with hyperthreading on your processor. Is it a quad-core?
It's a well known fact that you see a boost in flops when disabling HT...but I'm not sure if that would be a good idea or not in this case. (intel burn test surely likes that)
Does reducing the # of threads increase chainspermin for anyone else?
On my i7 I started using 6 threads and saw 8-12 CPMs.
When I switched to -1 it dropped to 3-6 CPMs.
I changed back to 6 and it increased again.
On my i7 I started using 6 threads and saw 8-12 CPMs.
When I switched to -1 it dropped to 3-6 CPMs.
I changed back to 6 and it increased again.
I wasn't able to replicate this. Maybe this has something to do with hyperthreading on your processor. Is it a quad-core?
difficulty 9 already, that's the reason
What's the reason it is only 9?
I'm trying to get my head around the idea that Quarkcoin launched a few hours ago and is already 57 difficultly.*
Is that an apples to oranges comparison? Surely Quarkcoin can't have 6 times the network strength of Primecoin.
As Primecoin difficulty rises, the supply of coins really gets choked off - at difficultly of 32 . . . that's 999/(32*32)= .975 coins per block.
*Oh, it's gone up to 63 since I wrote those two sentences.
It is very much like comparing apples and oranges. For Primecoin, the integer part of difficulty determines the required probable prime chain length. E.g. if difficulty is 9.07233632, a probable prime chain of length 9 is required. The fractional part is used to determine how many length-9 chains are accepted. It scales logarithmically so that going from 9.8 to 9.9 is a lot harder than going from 9.1 to 9.2. The details for the fractional part are in Sunny King's paper.
The current algorithm also does more work for longer target chain lengths. That's why PPS numbers are also dropping.
It's looking like mining is getting significantly less profitable without a gpu miner.
says the guy with 400-500k pps. lulz
If everyone had a GPU miner... you know how that would go.
Does reducing the # of threads increase chainspermin for anyone else?
On my i7 I started using 6 threads and saw 8-12 CPMs.
When I switched to -1 it dropped to 3-6 CPMs.
I changed back to 6 and it increased again.
On my i7 I started using 6 threads and saw 8-12 CPMs.
When I switched to -1 it dropped to 3-6 CPMs.
I changed back to 6 and it increased again.
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 for instructions.
make: *** [obj/alert.o] Error 4
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
make: *** [obj/alert.o] Error 4
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.
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.
It's looking like mining is getting significantly less profitable without a gpu miner.
says the guy with 400-500k pps. lulz
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.
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.
-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? 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.
I'm in PPS envy.
-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? 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.
-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? 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.
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.
-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? 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.
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??
Any fix??
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.
It's looking like mining is getting significantly less profitable without a gpu miner.
Jump to: