LZ
Legendary
Offline
Activity: 1722
Merit: 1072
P2P Cryptocurrency
|
 |
July 26, 2013, 01:02:05 AM |
|
IMPORTANT: Fixed loss of potential blocks caused by fractional length calculation being skipped if fast divisibility test succeeds. (Thanks to mtrlt for spotting this.) So it is fully fixed in the just released version hp8, right?  I might have found a bug in hp7. I copied the hp7 code over to Reaper, and found some weird behavior. I noticed that many shares/blocks it found, had a difficulty with a fractional part of 0.
Function: FermatProbablePrimalityTestFast. It does a Fermat test, and if it fails, it calculates the fractional part. However, there's a fast division test before the Fermat test. If the fast division test is succesful, the function is returned from and the fractional part isn't calculated. If it just happened to be the last number in a chain, the fractional part isn't calculated and is left at 0. This means if the difficulty is for example 6.2, and the miner found a block of difficulty 6.3, it's possible that the fractional part was left uncalculated, and the miner thinks it's difficulty 6.0. The block doesn't get submitted and lots of profit has been lost.
The same applies to EulerLagrangeLifchitzPrimalityTestFast.
Right now I don't have git set up, I can't submit a patch.
|
My OpenPGP fingerprint: 5099EB8C0F2E68C63B4ECBB9A9D0993E04143362
|
|
|
|
ReCat
|
 |
July 26, 2013, 01:17:02 AM |
|
Wow. Double?! That's pretty impressive, I was thinking the performance of the miner was already squeezing out every last drop.
|
BTC: 1recatirpHBjR9sxgabB3RDtM6TgntYUW Hold onto what you love with all your might, Because you can never know when - Oh. What you love is now gone.
|
|
|
rethaw
|
 |
July 26, 2013, 01:27:42 AM |
|
Wow. Double?! That's pretty impressive, I was thinking the performance of the miner was already squeezing out every last drop.
Better numbers, but haven't found any blocks!
|
|
|
|
eCoinomist
Member

Offline
Activity: 112
Merit: 10
Independent Analyst
|
 |
July 26, 2013, 02:33:18 AM |
|
Wow. Double?! That's pretty impressive, I was thinking the performance of the miner was already squeezing out every last drop.
Better numbers, but haven't found any blocks! I stopped mining for a few days due to recent clients crashing too much, just tried new version and found a block within 30 mins lol (using the same old simple compilation I had used since hp2)
|
|
|
|
Tuck Fheman
|
 |
July 26, 2013, 02:47:52 AM |
|
I stopped mining for a few days due to recent clients crashing too much, just tried new version and found a block within 30 mins lol (using the same old simple compilation I had used since hp2)
what are your settings if not default? "roundsievepercentage" : ??, "sievepercentage" : ??, "sievesize" : ??,
|
|
|
|
Dsfyu
Member

Offline
Activity: 75
Merit: 10
|
 |
July 26, 2013, 02:54:49 AM |
|
Thanks for the update - I'm seeing roughly a 40% increase in pps right now - averaging 6400 instead of about 4600 before. I'm still getting roughly 10-20 cpm with an average of about 15-17 (no change). Now lets see if this affects how often I find blocks...
|
Don't just trade, get paid to Atomic⚛Trade !!!Disclaimer: I am a noob. Assume I know nothing until proven otherwise.
|
|
|
AstroKev
Newbie
Offline
Activity: 23
Merit: 0
|
 |
July 26, 2013, 03:01:07 AM |
|
Seeing nearly 2x improvement. Amazing, great work. > ./primecoind getmininginfo { "blocks" : 83381, "chainspermin" : 58, "currentblocksize" : 1000, "currentblocktx" : 0, "difficulty" : 9.26739043, "errors" : "", "generate" : true, "genproclimit" : -1, "roundsievepercentage" : 30, "primespersec" : 33445, "pooledtx" : 0, "sievepercentage" : 10, "sievesize" : 100000, "testnet" : false }
|
|
|
|
Krusher33
Newbie
Offline
Activity: 37
Merit: 0
|
 |
July 26, 2013, 03:07:04 AM |
|
How does this compare to hp8 in Linux.
|
|
|
|
Dsfyu
Member

Offline
Activity: 75
Merit: 10
|
 |
July 26, 2013, 03:16:18 AM |
|
Seeing nearly 2x improvement. Amazing, great work. > ./primecoind getmininginfo { "blocks" : 83381, "chainspermin" : 58, "currentblocksize" : 1000, "currentblocktx" : 0, "difficulty" : 9.26739043, "errors" : "", "generate" : true, "genproclimit" : -1, "roundsievepercentage" : 30, "primespersec" : 33445, "pooledtx" : 0, "sievepercentage" : 10, "sievesize" : 100000, "testnet" : false }
What sort of hardware are you running this on right now? With my 3930k and those settings I get roughly 10k pps and anything from 10 to 40 cpm (still mostly at about 15)
|
Don't just trade, get paid to Atomic⚛Trade !!!Disclaimer: I am a noob. Assume I know nothing until proven otherwise.
|
|
|
Tuck Fheman
|
 |
July 26, 2013, 03:19:54 AM |
|
With my 3930k and those settings I get roughly 10k pps and anything from 10 to 40 cpm (still mostly at about 15)
Try ... "roundsievepercentage" : 20, "sievepercentage" : 8, "sievesize" : 100000,
|
|
|
|
eCoinomist
Member

Offline
Activity: 112
Merit: 10
Independent Analyst
|
 |
July 26, 2013, 03:30:50 AM |
|
I stopped mining for a few days due to recent clients crashing too much, just tried new version and found a block within 30 mins lol (using the same old simple compilation I had used since hp2)
what are your settings if not default? "roundsievepercentage" : ??, "sievepercentage" : ??, "sievesize" : ??, default everything.
|
|
|
|
Dsfyu
Member

Offline
Activity: 75
Merit: 10
|
 |
July 26, 2013, 03:32:49 AM |
|
Try ...
"roundsievepercentage" : 20, "sievepercentage" : 8, "sievesize" : 100000,
Thanks for the tip - I'm now getting 11k pps - I'm getting about 20 cpm now but it will randomly drop down to 10 cpm and then jump back up 23:32:42  { "blocks" : 83418, "chainspermin" : 21, "currentblocksize" : 1000, "currentblocktx" : 0, "difficulty" : 9.26809406, "errors" : "", "generate" : true, "genproclimit" : -1, "roundsievepercentage" : 20, "primespersec" : 11280, "pooledtx" : 0, "sievepercentage" : 8, "sievesize" : 100000, "testnet" : false }
|
Don't just trade, get paid to Atomic⚛Trade !!!Disclaimer: I am a noob. Assume I know nothing until proven otherwise.
|
|
|
gigawatt
|
 |
July 26, 2013, 03:39:29 AM |
|
What's better to optimize for: primes/sec, chains/min, or 5-chains/hr?
|
|
|
|
maco
|
 |
July 26, 2013, 03:39:52 AM |
|
IMPORTANT: Fixed loss of potential blocks caused by fractional length calculation being skipped if fast divisibility test succeeds. (Thanks to mtrlt for spotting this.) So it is fully fixed in the just released version hp8, right?  I might have found a bug in hp7. I copied the hp7 code over to Reaper, and found some weird behavior. I noticed that many shares/blocks it found, had a difficulty with a fractional part of 0.
Function: FermatProbablePrimalityTestFast. It does a Fermat test, and if it fails, it calculates the fractional part. However, there's a fast division test before the Fermat test. If the fast division test is succesful, the function is returned from and the fractional part isn't calculated. If it just happened to be the last number in a chain, the fractional part isn't calculated and is left at 0. This means if the difficulty is for example 6.2, and the miner found a block of difficulty 6.3, it's possible that the fractional part was left uncalculated, and the miner thinks it's difficulty 6.0. The block doesn't get submitted and lots of profit has been lost.
The same applies to EulerLagrangeLifchitzPrimalityTestFast.
Right now I don't have git set up, I can't submit a patch.
Let's get her done 
|
|
|
|
Tuck Fheman
|
 |
July 26, 2013, 04:02:27 AM |
|
but it will randomly drop down to 10 cpm and then jump back up
try setting roundsievepercentage : 40, it seemed to smooth results out for me with certain settings.
|
|
|
|
Dsfyu
Member

Offline
Activity: 75
Merit: 10
|
 |
July 26, 2013, 04:09:15 AM |
|
but it will randomly drop down to 10 cpm and then jump back up
try setting roundsievepercentage : 40, it seemed to smooth results out for me with certain settings. Right now it seems to be working fine and fairly stable with 30 - I'm just leaving the settings alone for now and just letting it mine So far I have gotten three blocks on my desktop (all on hp7) and one on my laptop (on hp8 now) just letting them sit and mine over the last three days - 1 block on 7/23 in the middle of the day, one on 7/24 in the early evening, one on 7/25 a little after midnight, and one about an hour ago on the laptop - seems fairly random as I haven't really played with my laptop settings yet and they definitely aren't optimized. Additionally all those I found on my desktop were on hp7 which meant they were effectively at a higher difficulty. I wonder how this will end up doing over the next day or two on hp8 - maybe we will find some even better settings that can get us a higher block rate.
|
Don't just trade, get paid to Atomic⚛Trade !!!Disclaimer: I am a noob. Assume I know nothing until proven otherwise.
|
|
|
Krusher33
Newbie
Offline
Activity: 37
Merit: 0
|
 |
July 26, 2013, 04:12:59 AM Last edit: July 26, 2013, 04:26:48 AM by Krusher33 |
|
Thank you mikaelh! "blocks" : 83454, "chainspermin" : 13, "currentblocksize" : 1191, "currentblocktx" : 1, "difficulty" : 9.26784617, "errors" : "", "generate" : true, "genproclimit" : 7, "roundsievepercentage" : 20, "primespersec" : 10040, "pooledtx" : 1, "sievepercentage" : 8, "sievesize" : 100000, "testnet" : false
I just wish I could get my chainspermin up. tried sievepercentage 6 and 12, and tried roundsievepercentage 40... they did not do better than the above.
|
|
|
|
Tuck Fheman
|
 |
July 26, 2013, 06:16:09 AM |
|
This is a testnet result, but wow ...
{ "blocks" : 58199, "chainspermin" : 52, <--- my highest by 3x "currentblocksize" : 1000, "currentblocktx" : 0, "difficulty" : 6.30837959, "errors" : "", "generate" : true, "genproclimit" : -1, "roundsievepercentage" : 30, "primespersec" : 6210, <--- my highest "pooledtx" : 0, "sievepercentage" : 10, "sievesize" : 9400000, <--- did no good on mainnet. slow pps, 0-1 CPM. "testnet" : true }
I never imagined I'd see 52 CPM, even on testnet. I've never seen above 16.
Higher sievesize works better on testnet, lower sievesize works better (higher PPS/CPM) on mainnet.
|
|
|
|
Tuck Fheman
|
 |
July 26, 2013, 06:19:29 AM |
|
I just wish I could get my chainspermin up. tried sievepercentage 6 and 12, and tried roundsievepercentage 40... they did not do better than the above.
did you try 100K/10/30? Here are the settings that yielded the highest CPM w/good to great PPS. "roundsievepercentage" : 40, "sievepercentage" : 8, "sievesize" : 200000, "roundsievepercentage" : 25, "sievepercentage" : 10, "sievesize" : 100000, "roundsievepercentage" : 20, "sievepercentage" : 8, "sievesize" : 100000, "roundsievepercentage" : 30, "sievepercentage" : 10, "sievesize" : 100000,
|
|
|
|
|