Bitcoin Forum
May 27, 2024, 07:13:14 AM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 [17] 18 19 »
321  Bitcoin / Mining / My theory of shares by volume vs. speed on: February 27, 2012, 08:20:04 AM
Please don't judge on my math: (irony: I'm a math minor and never taken statistics)

Well here goes:

Given: 1 share = 1/(2^32) hashes

Group A:
10 GPU's producing 100 Mh/s each
Total: 1 Gh/s

Group B:
1 GPU Producing 1 Gh/s*
Total: 1 Gh/s

*Disclamer: THIS IS THEORY ONLY

Prove: Mining in a proportional pool, you will benefit more from one gpu providing maximum hashing speed, vs multiple cards hashing at slower speeds.

A:
10 Gpu's mining @ 100 Mh/s simultaneously will provide roughly 1 share every 42.9 seconds.
((2^32) hashes / 1 share) * (1 s / 100E6 hashes) = 42.94 s/share

B: 1 Gpu mining @ 1 Gh/s will provide roughly 1 share every 4.29 seconds
((2^32) hashes / 1 share) * (1 s / 1E9 hashes) = 4.29 s/share


ALL POOLS USE A NON-FEE PROPORTIONAL METHOD
Pool has a constant 1 Th/s collective speed.
Scenario 1: (extreme case)
Pool finds a block in 30 seconds.
Total shares: (1 share / (2^32) hashes) * (1E12 hashes / 1 s) * 30 s = 6,980 shares (3 sig figs)

A:
Shares by you: (1 share / 42.9 s) * 30 s = 0.699 shares / GPU = 0 shares
Payout: ( 0 shares / 6980 shares ) * 50 BTC = 0 BTC

B:
Shares by you: (1 share / 4.29 s)  * 30 s = 6.99 shares (tough luck) = 6 shares
Payout: ( 6 shares / 6980 shares ) * 50 BTC = 0.430 BTC

Scenario 2:
Pool finds a block in 180 seconds.
Total shares: (1 share / (2^32) hashes) * (1E12 hashes / 1 s) * 180 s = 41,900 shares (3 sig figs)

A:
Shares by you: (1 share / 42.9 s) * 180 s = 4.196 shares / GPU = 40 shares
Payout: ( 40 shares / 41900 shares ) * 50 BTC = 0.477 BTC

B:
Shares by you: (1 share / 4.29 s) * 180 s= 41.95 shares / GPU = 41 shares
Payout: ( 41 shares / 41900 shares ) * 50 BTC = 0.489 BTC

Scenario 3:
Pool finds a block in 3600 seconds.
Total shares: (1 share / (2^32) hashes) * (1E12 hashes / 1 s) * 3600 s =  838,000 shares (3 sig figs)

A:
Shares by you: (1 share / 42.9 s) * 3600 s = 83.9 shares / GPU = 830 shares
Payout: ( 830 shares / 838000 shares ) * 50 BTC = 0.495 BTC

B:
Shares by you: (1 share / 4.29 s) * 3600 s = 839 shares / GPU = 839 shares
Payout: ( 839 shares / 838000 shares ) * 50 BTC = 0.506 BTC

In conclusion:

Stated by my data, the highest hash rate obtainable by one card vs multiple smaller cards will almost always yield more BTC over time.

DISCLAIMER:  This is a lottery.  You can get get a share the first hash you do with any card.  This data represents results over a large period of time.

This is a picture of the results in an excel spreadsheet:



This can be related to 5870's vs 5970's.

~400 Mh/s to ~350 Mh/s
This theory is also in favor of the BFL 850 Mh/s miner.  Go figure.
322  Bitcoin / Pools / Re: [3600 Gh/s] DeepBit.net PPS+Prop,instant payouts, we pay for INVALID BLOCKS too on: February 24, 2012, 08:38:13 PM
Gah Deepbit whats wrong with you?
Code:
 cgminer version 2.1.2 - Started: [2012-02-22 10:49:29]
--------------------------------------------------------------------------------
 (5s):331.2 (avg):609.3 Mh/s | Q:28505  A:25074  R:64  HW:0  E:88%  U:8.43/m
 TQ: 1  ST: 0  SS: 51  DW: 1363  NB: 332  LW: 0  GF: 112  RF: 149
 Connected to http://pit.deepbit.net:8332 with LP as user ___________
 Block: 000001b45780cb1cb1d5f66ee975187f...  Started: [12:20:59]
--------------------------------------------------------------------------------
 [P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
 GPU 0:  64.0C  40%    | 314.3/304.2Mh/s | A:12581 R:32 HW:0 U:  4.23/m I: 3
 GPU 1:  65.0C  40%    | 439.1/305.2Mh/s | A:12493 R:32 HW:0 U:  4.20/m I: 3
--------------------------------------------------------------------------------

[2012-02-24 12:24:18] Pool 0 not providing work fast enough
[2012-02-24 12:24:53] Pool 0 communication resumed, submitting work
[2012-02-24 12:24:53] Accepted 00000000.4e68a616.2f23a60a GPU 1 thread 1
[2012-02-24 12:24:53] Accepted 00000000.063b83c7.84ad558f GPU 0 thread 2
[2012-02-24 12:24:53] Accepted 00000000.c7960201.c8289535 GPU 0 thread 0
[2012-02-24 12:24:54] Accepted 00000000.6b58e6e8.b8ff2c1d GPU 1 thread 1
[2012-02-24 12:24:56] Pool 0 communication failure, caching submissions
[2012-02-24 12:24:57] Pool 0 communication resumed, submitting work
[2012-02-24 12:24:57] Accepted 00000000.7348486d.dd80b66c GPU 0 thread 2
[2012-02-24 12:25:07] Pool 0 not providing work fast enough
[2012-02-24 12:25:19] Pool 0 http://pit.deepbit.net:8332 not responding!
[2012-02-24 12:25:19] Pool 0 communication failure, caching submissions

Edit: It was prolly my fault

I'm having the same problems.  I'm at like.. 25% rejection rate on my rig.
323  Other / CPU/GPU Bitcoin mining hardware / Minimizing Rejects on: February 24, 2012, 06:12:46 AM
I had my 5970 running at about 0.1% rejects for ~ 48 hours.

When I added my 5850 into the mix, my rejects on all cards are up to about 0.5%

What are the most effective ways to minimize rejects?

I've tried playing with my voltage and it seems to have no effect on it.
324  Other / CPU/GPU Bitcoin mining hardware / Re: Compiled Undervolting Results on: February 24, 2012, 05:39:35 AM
Miner: pheonix
OC software: bamt
Card: 5970
Core: 775
Memory: 300
Voltage: 1.02
Mh/s: 704

Miner: pheonix
OC software: bamt
Card: 5850
Core: 775
Memory: 300
Voltage: 1.02
Mh/s: 341
325  Other / CPU/GPU Bitcoin mining hardware / Re: Unofficial "High MHash"/"Best Settings" GPU Comparision Thread on: February 24, 2012, 05:31:06 AM
Please post only if you have beaten the previous MH/s record for that card.

Going to reiterate myself a bit.  Wink

Thanks for the 5970 & 6870 results nonetheless.

I can't agree with that.  I'm more comfortable running my cards cooler and with less voltage.  Less Power = more profit

Edit:  It would be nice if you made a separate table for alternate configurations.
326  Other / CPU/GPU Bitcoin mining hardware / Re: Unofficial "High MHash"/"Best Settings" GPU Comparision Thread on: February 24, 2012, 04:22:55 AM
Card Manufacturer: ATI
Card Model Number: 5970
Card OEM: Powercolor
MHash per Second: 352/352
Core Clock: 775/775
Vddc: 1.02/1.02V
Memory Clock: 300/300
Average Operating Temperature: 65/67C
Ambient Temperature: 70F
Fan Speed: 50
Host OS: BAMT 0.4b
Driver Version: SDK 2.4
Mining Program: Pheonix
Command Line Flags/GUIminer Settings: default

Card Manufacturer: ATI
Card Model Number: 5850
Card OEM: Asus
MHash per Second: 341
Core Clock: 825
Vddc: 1.02
Memory Clock: 300
Average Operating Temperature: 62C
Ambient Temperature: 70F
Fan Speed: 35
Host OS: BAMT 0.4b
Driver Version: SDK 2.4
Mining Program: Pheonix
Command Line Flags/GUIminer Settings: default

Card Manufacturer: ATI
Card Model Number: 5850
Card OEM: Sapphire (non reference)
MHash per Second: 347
Core Clock: 825
Vddc: 1.02
Memory Clock: 300
Average Operating Temperature: 64C
Ambient Temperature: 70F
Fan Speed: 62
Host OS: BAMT 0.4b
Driver Version: SDK 2.4
Mining Program: Pheonix
Command Line Flags/GUIminer Settings: default

Card Manufacturer: ATI
Card Model Number: 6870
Card OEM: Sapphire
MHash per Second: 271
Core Clock: 950
Vddc: 1.175
Memory Clock: 525
Average Operating Temperature: 70C
Ambient Temperature: 70F
Fan Speed: 60
Host OS: Windows 7
Driver Version: SDK 2.4
Mining Program: Cgminer
Command Line Flags/GUIminer Settings: -v 2 -w 128
327  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 23, 2012, 12:30:58 AM
Underscores don't work either.  I changed mine to a (-) and it works now.
328  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 22, 2012, 04:58:33 AM
It is running.  That's good I think.  Is there a config option to tell it to monitor?


I tried to cycle it. I will report back with results.

edit: Still no luck, but


It just stays like that.  It hasn't worked since I updated bamt.  Before that it did.  Maybe it was something with the munin update?

So, is there anyway to redo an update?



So, according to this, it can't find my plugins? ><
329  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 22, 2012, 04:50:47 AM
the /munin directory is a vdir to /tmp/munin/html/

so, if you: ls /tmp/munin/html/

you should see there a directory rig_1, and inside an index.html

do you?

if so.. did you break the vdir? maybe when switching the port?

if not, munin-update or one of its sub processes isn't working properly.  look at /tmp/munin/log/* for clues.




Nothing there :/

All I have is 3 "connection timed out" warnings in the log.  Is there a command to re-install munin with the same settings?
330  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 22, 2012, 04:39:25 AM
is your miner_id set to something other than exactly 'rig_1' ?


Nope, exactly that.

edit:


I poked around in the directory and found that
331  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 22, 2012, 04:27:49 AM


This has been going on for awhile now, and I've decided to try and fix it.  The graphs used to work until I updated BAMT.  The hashrate graphs don't show on the status.pl page either. I've searched for a few hours on google and can't seem to find anything relating to the error.  Anyone know what to do?
332  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 11:34:29 PM
Neat, thanks. I'll play around with it.  Learning more about linux every day Cheesy  This is what makes BAMT so great; the awesome support behind it Cheesy
333  Other / CPU/GPU Bitcoin mining hardware / Re: Radeon 5850 stuck at 75-80% on: February 21, 2012, 11:32:48 PM
Solved! thanks!
334  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 11:31:05 PM


Thank you very much!

That's what i get for typing it on SSH.  Is there an easy way to copy/paste with vi?
335  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 11:25:51 PM
removed Smiley thank you
336  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 11:22:49 PM
I'm not sure weather I'm doing this right, but I don't get any params.  I just get a bunch of sh, pheonix.py and bash returned

337  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 11:16:19 PM
-Stopped all miners

-Started two ssh connections

-wrapper 2

-top

showing 25% cpu usage at 88% gpu usage underclocked to 350mhz

verify clocks now
338  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 11:06:18 PM


underclocked to 350mhz and got 88% :/

Do I have a bad GPU?
339  Other / CPU/GPU Bitcoin mining hardware / Re: Radeon 5850 stuck at 75-80% on: February 21, 2012, 10:54:33 PM
did you re-initialize X after changing (adding) the card configuration?

sudo aticonfig -f --initial --adapter=all ; sudo coldreboot

what does aticonfig tell you about the clocks on the card, sometimes bamt gets confused I've found

sudo aticonfig --od-enable ; export DISPLAY=:0 ; sudo aticonfig --odgc --adapter=all

774MHz seems low for a 5870.  I think the stock clocks are 850 or 900 depending on the brand.  One of my miners is the exact same config as you described...FWIW, I run at 835/300 on the 5970 and 950/300 on the 5870 and pull 1212MH/s all day long from that miner

********************
edit:  just saw you have a 5850, not 5870...my bad...do check the clocks though bamt can/does get confused from time to time.

I did not reinitialize it.  I didn't know I had to do anything besides plug and play.  Rebooting now from first step to see if it changes
340  Bitcoin / Project Development / Re: BAMT - Easy persistent USB key based linux for dedicated miners/mining farms on: February 21, 2012, 09:37:01 PM


So I got a new 5850 today, and my load is staying at ~80%

I copied the config over and removed all overclocking besides memclock.  What am I doing wrong? o.o

Update on my situation:

Found out the 5850 is using 50% cpu (two cores)

My box is 100% updated.  Has anyone had this bug before?
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 [17] 18 19 »
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!