Mikestang
Legendary
Offline
Activity: 1274
Merit: 1000
|
|
October 07, 2015, 05:02:19 AM |
|
sidehack sticks do 55% of freq
I think you mean 5.5% of frequency? Checking the math it comes out to 5.5 at least.
|
|
|
|
TheRealSteve
|
|
October 07, 2015, 11:21:26 AM |
|
Phil dentally picked the right solo pool to use . Dentally? Well I suppose he did sink his teeth into it I initially suggested it because of the relative ease of adding workers, rather than having to assign individual addresses etc. Helps that ck (and kano) are very responsive
I think you mean 5.5% of frequency? Checking the math it comes out to 5.5 at least.
The ASIC's clock -> Gh/s multiplier is 55. 100MHz * 55 = 5500MHz -> 5.5Gh/s So 5.5% is correct..albeit off by a factor 1,000
I was thinking it might also be fun to keep track of who currently has the "best share" reported by the club's address. As of this post the best reported share is 51,471,144. This belongs to philipma1957donation. Strange, though, because FinksySticks shows a higher best share of 65 million, what gives? Maybe the /users/ one gets reset, but /worker/ does not? ck might have to answer that. Right now FinksySticks doesn't show nearly that high of a best share. The bot (ignoring resets) reports: Top club best shares: valkirstick (891,953,402) philipma1957sticks (216,364,341) fullzerostick (191,610,039) FinksySticks (65,558,032) bmoscatosticks (49,883,780) Top donator best shares: philipma1957donation (2,366,037,588) philipma1957s7 (1,305,528,734) slimepuppydonation (176,899,661) aarons6 (82,192,528) larry12s5donation (49,990,406)
|
|
|
|
toptek
Legendary
Offline
Activity: 1274
Merit: 1000
|
|
October 07, 2015, 11:52:34 AM |
|
okay here are 20 sticks at freq 250 for more then 23 hours one iffy by iffy 100 hw out of 82000 which is under .12 percent 19 perfect not bad. the stud hub on the left has the best 5volt psu for efficiency (90% rating rated for 150watts with no cooling 200 watts with cooling) but it has to be outside the stud hub. So I pulled the board and built a platform to mount it all. I am running 15 sticks the stud hub with 5 sticks has an 85% psu rated for 130 watts temps are 81 f power for the 2 hubs = 128watts in one of your Picture not shown here i saw a USB thingie with digital numbers on it what was that i want one.
|
|
|
|
baktegon
Newbie
Offline
Activity: 27
Merit: 0
|
|
October 07, 2015, 01:04:59 PM |
|
Guys I'm being a dummy. I'm looking for the chart that compares frequency to GH but I can't find it. Even in the Google! If someone has the link i'd appreciate it.
sidehack sticks do 55% of freq so 125 freq = 6.875 gh 150 freq = 8.250 gh 175 freq = 9.625 gh 200 freq = 11.000 gh 225 freq = 12.375 gh 250 freq = 13.750 gh 275 freq = 14.135 gh 300 freq = 16.500 gh 325 freq = 17.875 gh 350 freq = 19.250 gh 375 freq = 20.625 gh 400 freq = 22.000 gh I run 20 sticks at freq 250 I should do 275gh I am doing about 272gh Oh perfect, saved it for later! Thanks Phil!
|
|
|
|
|
|
Mikestang
Legendary
Offline
Activity: 1274
Merit: 1000
|
|
October 07, 2015, 04:44:31 PM |
|
I think you mean 5.5% of frequency? Checking the math it comes out to 5.5 at least.
The ASIC's clock -> Gh/s multiplier is 55. 100MHz * 55 = 5500MHz -> 5.5Gh/s So 5.5% is correct..albeit off by a factor 1,000 You're adding an extra step or two to make 55% work, no need to mess around with 1000MHz = 1GH/s conversion like you did.. If you look at the list Phil posted earlier, which I've quoted and editted below, you take the frequency and multiply by 0.055 (5.5%) you get the hash in GH/s. 125 freq * 0.055 = 6.875 gh 150 freq * 0.055 = 8.250 gh 175 freq * 0.055 = 9.625 gh 200 freq * 0.055 = 11.000 gh 225 freq * 0.055 = 12.375 gh 250 freq * 0.055 = 13.750 gh 275 freq * 0.055 = 14.135 gh etc.
I was thinking it might also be fun to keep track of who currently has the "best share" reported by the club's address. As of this post the best reported share is 51,471,144. This belongs to philipma1957donation. Strange, though, because FinksySticks shows a higher best share of 65 million, what gives? Maybe the /users/ one gets reset, but /worker/ does not? ck might have to answer that. Right now FinksySticks doesn't show nearly that high of a best share. Yes you're right. It wasn't intentional, but an oversight on my part that workers' best share doesn't reset along with the rest. Might not be such a bad thing...
It's not showing that high of a share because it reset at some point, probably after a solo solve yesterday? But if ck says worker best share does not reset then I don't know how else to explain the discrepancy.
|
|
|
|
pepto
|
|
October 07, 2015, 06:38:52 PM |
|
The only concern I have with resetting one's best share with each block solve is that I will never know if luck is a true thing. If my sticks never solve anything ever for years, yea I know, come on what are you expecting, if I saw a best share of mine that was even in the ballpark I would hope that you would share my excitement.
|
|
|
|
edonkey
Legendary
Offline
Activity: 1150
Merit: 1004
|
|
October 07, 2015, 07:08:01 PM |
|
The only concern I have with resetting one's best share with each block solve is that I will never know if luck is a true thing. If my sticks never solve anything ever for years, yea I know, come on what are you expecting, if I saw a best share of mine that was even in the ballpark I would hope that you would share my excitement.
I used to feel the same way. Even through best share is basically a meaningless statistic, it does give you a sense that something is happening. By default, my monitoring script will sent an email if any of the monitored addresses have an improved best share. The last best share is stored in a file locally, so it doesn't matter if the pool resets it or not. When I was just monitoring my own addresses, this was cool because I'd get an email once in a while. But now that I monitor all of the club's addresses, I was getting a ton of best share emails. It was really too much so I disabled monitoring of best shares for now. Now all I see is when someone (anyone) solves a block on CK's solo pool. So far it's always been someone else. But where there's hash, there's hope
|
Was I helpful? BTC: 3G1Ubof5u8K9iJkM8We2f3amYZgGVdvpHr
|
|
|
TheRealSteve
|
|
October 07, 2015, 08:07:33 PM |
|
The ASIC's clock -> Gh/s multiplier is 55. 100MHz * 55 = 5500MHz -> 5.5Gh/s So 5.5% is correct..albeit off by a factor 1,000 You're adding an extra step or two to make 55% work, no need to mess around with 1000MHz = 1GH/s conversion like you did.. If you look at the list Phil posted earlier, which I've quoted and editted below, you take the frequency and multiply by 0.055 (5.5%) you get the hash in GH/s. I think you're mildly confused Phil's list is fine, but mentioned 55%, which is incorrect. I said "multiplier is 55", that's not a percentage 100MHz * 55h/Hz = 5500Mh/s or 5.5Gh/s. No extra step there, and no oddball percentage "divide by 100, multiply by 5.5 (aka percentages), then go from Mega to Giga by just substituting strings"
|
|
|
|
toptek
Legendary
Offline
Activity: 1274
Merit: 1000
|
|
October 07, 2015, 09:11:04 PM |
|
|
|
|
|
-ck
Legendary
Offline
Activity: 4284
Merit: 1645
Ruu \o/
|
|
October 07, 2015, 10:16:41 PM |
|
Maybe the /users/ one gets reset, but /worker/ does not? ck might have to answer that. Right now FinksySticks doesn't show nearly that high of a best share.
Yes you're right. It wasn't intentional, but an oversight on my part that workers' best share doesn't reset along with the rest. Might not be such a bad thing... EDIT: No, the worker best share does reset, as that code was there from before. It's only the share count that never resets, and that applies to users and workers.
|
Developer/maintainer for cgminer, ckpool/ckproxy, and the -ck kernel 2% Fee Solo mining at solo.ckpool.org -ck
|
|
|
philipma1957 (OP)
Legendary
Offline
Activity: 4298
Merit: 8768
'The right to privacy matters'
|
|
October 07, 2015, 10:17:19 PM |
|
I do .55 x 125 = 68.75 why? it is the least amount of keys to enter on a calculator. I then type in freq = 6.875gh in the chart. just moving the decimal place cause I know a stick cant do 68.75 Technically entering .055 x 125 = 6.875 is correct but it is not faster and the chart has 10 plus entries. My bad since I did not explain it exactly. I think you mean 5.5% of frequency? Checking the math it comes out to 5.5 at least.
The ASIC's clock -> Gh/s multiplier is 55. 100MHz * 55 = 5500MHz -> 5.5Gh/s So 5.5% is correct..albeit off by a factor 1,000 You're adding an extra step or two to make 55% work, no need to mess around with 1000MHz = 1GH/s conversion like you did.. If you look at the list Phil posted earlier, which I've quoted and editted below, you take the frequency and multiply by 0.055 (5.5%) you get the hash in GH/s. 125 freq * 0.055 = 6.875 gh 150 freq * 0.055 = 8.250 gh 175 freq * 0.055 = 9.625 gh 200 freq * 0.055 = 11.000 gh 225 freq * 0.055 = 12.375 gh 250 freq * 0.055 = 13.750 gh 275 freq * 0.055 = 14.135 gh etc.
I was thinking it might also be fun to keep track of who currently has the "best share" reported by the club's address. As of this post the best reported share is 51,471,144. This belongs to philipma1957donation. Strange, though, because FinksySticks shows a higher best share of 65 million, what gives? Maybe the /users/ one gets reset, but /worker/ does not? ck might have to answer that. Right now FinksySticks doesn't show nearly that high of a best share. Yes you're right. It wasn't intentional, but an oversight on my part that workers' best share doesn't reset along with the rest. Might not be such a bad thing...
It's not showing that high of a share because it reset at some point, probably after a solo solve yesterday? But if ck says worker best share does not reset then I don't know how else to explain the discrepancy.
|
|
|
|
Mikestang
Legendary
Offline
Activity: 1274
Merit: 1000
|
|
October 07, 2015, 10:18:49 PM |
|
I think you're mildly confused Phil's list is fine, but mentioned 55%, which is incorrect. I said "multiplier is 55", that's not a percentage Well, I guess you could call it confused if you mean I read what someone wrote and then applied what they wrote and determined they were wrong. So yes, now I see that Phil's post was in error with the "%" mention, that clears it up, carry on.
|
|
|
|
notlist3d
Legendary
Offline
Activity: 1456
Merit: 1000
|
|
October 07, 2015, 11:34:28 PM |
|
Mine is back up. Appears electricity was lost on RPI. It rebooted but sit there with no miners. So my average is lower today but should be fixed now.
Now need to find a scrypt to do run it automatically so does not happen again.
|
|
|
|
Jake36
|
|
October 08, 2015, 12:00:05 AM |
|
Mine is back up. Appears electricity was lost on RPI. It rebooted but sit there with no miners. So my average is lower today but should be fixed now.
Now need to find a scrypt to do run it automatically so does not happen again.
Easiest way is edit the /etc/rc.local (if your using raspbian) and add (just above the exit 0) cd /home/pi "or whatever you used"/cgminer-gekko sudo ./cgminer -o pool -u user --compac-freq xxx what ever else you want And it'll start mining as soon as it boot's.
|
|
|
|
edonkey
Legendary
Offline
Activity: 1150
Merit: 1004
|
|
October 08, 2015, 12:31:07 AM |
|
As an alternative, Minera automatically starts your miner. You can optionally specify a number of seconds to delay after boot, if that's interesting. It also supports automatically restarting if the miner software crashes (which I've never seen with cgminer-gekko), or restarting if N number of units goes non-responding (again, never needed to use this with the sticks). Minera also has a pretty GUI, if you like that
|
Was I helpful? BTC: 3G1Ubof5u8K9iJkM8We2f3amYZgGVdvpHr
|
|
|
pepto
|
|
October 08, 2015, 12:38:32 AM |
|
Mine is back up. Appears electricity was lost on RPI. It rebooted but sit there with no miners. So my average is lower today but should be fixed now.
Now need to find a scrypt to do run it automatically so does not happen again.
I don't know that this addresses what you're saying. I recently had a 20 minute whole house power failure. When it came back on, I deliberately did nothing just to see what my RPI Cgminer-Gekko setup would do. The pi restarted, the sticks restarted, minera restarted (or never went away ). It may have taken another twenty minutes for things to get back to where they were, but they did. And probably would have done it faster if I put a suggested share difficulty in there instead of waiting for that first 1K to get things moving. From someone who has never used a RPI, I am immensely impressed with it's ability to free me from my pc while mining 24/7. Of course we do the same thing with antminers and such when our blood is running hot, but for sticks this cannot be beat. And you edonkey are going to remain a hero to me for some time for your iteration of Minera/Cgminer-Gekko. It's absolutely sweet.
|
|
|
|
philipma1957 (OP)
Legendary
Offline
Activity: 4298
Merit: 8768
'The right to privacy matters'
|
|
October 08, 2015, 02:07:56 AM |
|
|
|
|
|
pepto
|
|
October 08, 2015, 03:03:28 AM |
|
I just slipped in 2 or so Gh/s to my stickname (maybe a high estimate) because i wanted to run my drillbit thumb for a while. Ha, how many of you know what a drillbit thumb is. I trust you understand I'm not asking how many of you don't give a darn, I can figure that for myself.
|
|
|
|
|