Bitcoin Forum
December 12, 2024, 09:41:48 AM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: Is it possible to get an accruate Mhash rate based on number of shares  (Read 1910 times)
Xenland (OP)
Legendary
*
Offline Offline

Activity: 980
Merit: 1003


I'm not just any shaman, I'm a Sha256man


View Profile
June 08, 2011, 12:49:18 AM
 #1

Believe me when I tell you, I know it sounds like a stupid question but seriously is it possible to get an accurate Mhash rate based on the share data pushpoold submits?
All my formulas come out accurate at first but once there its more or less workers they go off the charts in accurate!
I've seen other pooling services make this mistake of inaccuate hashing scores mine are usually at a ~20 Mhash/s error rate, which might piss some
new combing miners off when they come to find that they "Might" be making less Mhashes then their minors are reporting.

So is it possible to get an accruate mhash rate with the following data per share submitted
Fields:
id | timestamp | username

This is the last formula I attempted to use

Most recent share  submitted in time stamp
Minus(-)
First share submitted in round time stamp
Equals = Total Time Connected

TotalTimeConnected
Divided By(/)
Total Shares Subbmitted this round
Equals = Failed horribly at accurate Mhash Rate.
redhatzero
Full Member
***
Offline Offline

Activity: 126
Merit: 100



View Profile
June 08, 2011, 02:20:05 AM
 #2

I'm using s.th. Like this (got some hints in the Pushpool thread):

NrOfShares / timeframe * 4294967296

That should be hashes per second if pushpool is rewriting difficulty to 1

But due to the nature of shares, I wouldn't call that accurate. It will always be more like a guess


Xenland (OP)
Legendary
*
Offline Offline

Activity: 980
Merit: 1003


I'm not just any shaman, I'm a Sha256man


View Profile
June 08, 2011, 02:37:26 AM
 #3

I'm using s.th. Like this (got some hints in the Pushpool thread):

NrOfShares / timeframe * 4294967296

That should be hashes per second if pushpool is rewriting difficulty to 1

But due to the nature of shares, I wouldn't call that accurate. It will always be more like a guess



Is this the current difficulty?
4294967296
PwrLeveld
Member
**
Offline Offline

Activity: 336
Merit: 10


Computta Mine Your Own BTC


View Profile
June 08, 2011, 02:43:14 AM
 #4

No, I don't beleive so. However, I do not know the current dificulty but it says it in the wiki.

Dear GOD/GODS and/or MEMBERS OF SUPER-INTELLIGENT ALIEN CIVILIZATIONS: Please let CHAUL JHIN KIM win the NEW ZEALAND POWERBALL JACKPOT. He will use the money to set up his own RESEARCH LAB and find the CURES for CANCER and AGING! He will release the CURES into the PUBLIC DOMAIN! (a la Jonas Salk) EVERYONE regardless of wealth will no longer need to SUFFER! CHILDHOOD CANCER will also be HISTORY! Thank you! - A Faithful Believer. P.S. If anyone is reading this then please pray with me!
redhatzero
Full Member
***
Offline Offline

Activity: 126
Merit: 100



View Profile
June 08, 2011, 02:49:17 AM
 #5

No, it's 2^32 which should be the whole keyspace for one workpacket if I remember right

http://forum.bitcoin.org/index.php?topic=8707.msg137869#msg137869

M2Ys4U
Newbie
*
Offline Offline

Activity: 6
Merit: 0


View Profile
June 08, 2011, 04:42:14 AM
 #6

Hashes per second = Number of shares * 2^32 / time delta

Where time delta is the number of seconds you're reporting the hashrate average over.
Xenland (OP)
Legendary
*
Offline Offline

Activity: 980
Merit: 1003


I'm not just any shaman, I'm a Sha256man


View Profile
June 08, 2011, 08:43:36 AM
Last edit: June 08, 2011, 05:04:32 PM by Xenland
 #7

So i used this formula and i got 8 hashes per second?

Update: I figured out i did the formula backwards but its still not even close?

$hashesPerSecond =  $numShares / $fiveMinutesAgo * 4294967296;
The result is 16.345 Hahses/s
but my miner runs as 85000 Khash/s

UPDATE: I was using $fiveminutesago, and not the amount of time that has passed
M2Ys4U
Newbie
*
Offline Offline

Activity: 6
Merit: 0


View Profile
June 09, 2011, 01:01:01 AM
 #8

It will vary wildly. Using that formula, a 600 second time delta and approx 66MHash/s I get reports between 20MHash/s and 120MHash/s, but that's just statistical variance.
Xenland (OP)
Legendary
*
Offline Offline

Activity: 980
Merit: 1003


I'm not just any shaman, I'm a Sha256man


View Profile
June 09, 2011, 02:34:09 AM
 #9

It almost seems like we should use a Minutes Per Share graph instead of an Mhash graph since they are so unpredictably inaccurate.
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!