Bitcoin Forum
May 23, 2024, 10:43:42 PM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: [1] 2 »
1  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: July 12, 2013, 09:33:27 AM
Wow... I just barely made enough to cover electricity in the last 24 hrs and I'm running FPGAs and a Raspberry Pi.  I need to get solar power or something.
2  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: July 07, 2013, 10:46:07 AM
In fact, you are punished for disconnection. You're punished exactly the same way you'd been in PPS in the long run, provided that you disconnect asynchronously, i.e. randomly.
At Slush's, you can either lose, when you disconnect near the round end, or you can gain, when you disconnect at some other time. In the latter case, your payout for the round won't be decreased due to the disconnection, although you would deserve some decrease. And as the time constant c is hopefully still 300 seconds and an average round is several times longer, it's greater chance for you to gain a bit than to lose substantial part of the reward.
Anyway, in the long run that should even out, as is the popular phrase in this thread for some time already... :-)

(...)
but I have a hard time feeling I'm being punished for my downtime when ... well ... I'm down.

In the long run, you're "punished" just to the extent you're down.
When you won't come to your work, you won't be paid.
Or would you like something like paid holidays? Smiley

I swear this topic comes up every now and then because... well like you said, it sucks to be punished (e.g. kicked when you're down).

I want to elaborate a bit and you guys can correct me if I'm wrong.  The key to all this is really variance.  The way scoring is done on Slush's pool is subject to variance and it's much more evident on slower miners and less so (but still present) on faster/more powerful ones.  So what does variance have to do with being punished???... a lot.

Take for example if you're disconnected in the beginning of one of those crazy 7 hour rounds and 1 hour later you manage to fix your router, USB hub, whatever... At the end of that round you're actually not penalized that hard because what's 1 hour worth of shares compared to 6 hours of shares (and plus your score is isn't hit badly because you've been continuously mining for the last 6 hours of the round).

Now take the counter example.  You disconnect for 1 hour (same amount of time as before) but instead 4 rounds go by (a 5 min, 30 min, 15 min, and 10 min).  You just missed out on 4 different payouts.  Remember, on Slush you're paid based on the rate you contribute, so you usually get the same pay for large rounds as you do for short ones (again with some variance).

Contrast the two examples and what you get is HUGE variance in punishment.  You can get punished a little bit  or a lot.  While it's true that it averages out in the long run, the "long run" is a really really long time.  For someone who isn't down that often... say only once a month at most... you catch one bad break and you'll need at least 2, 3, or 4 lucky breaks to "even out".  I figure that most people don't catch as many lucky breaks as they do bad lucks as they haven't really been mining all that long (I speak from experience as both of the above scenarios have happened to me, plus I'm a low hash rate miner and I haven't mined that long).  You really can't compare this to something like PPS because they're two different scoring mechanism.  It's like comparing apples and oranges.  PPS just doesn't punish you the way Slush's scoring does.

So.... Why stay on Slush?  The reason is the same. While you get heavily punished for disconnections during short rounds, you can also get highly rewarded in short rounds (if you're always connected and mining).  As a slower/weaker hashing miner you get paid the same in short rounds as you do in long rounds (again with a little variance) and with lots of short rounds, you get paid more on Slush than if you're being paid PPS in the same period (partially due to the high pool fees associated with PPS pools). 

Really all that's left now if for someone to do some really hard number crunching.  Basically someone needs to look at the current difficulty, the percentage of Slush's pool hashing power as compared to the rest of the pools (and remaining hashing power not attributed to a pool) and generate a standard deviation curve for round duration.  Using the round duration curve against someone's hashing power, one can calculate their payout probability curve (per a given duration, say 24hours).  You can take that curve and mark that against what other pool's PPS rates are and then determine if the probability of earning a greater amount on Slush's pool vs other pool's PPS rate is statistically worth staying on Slush.
3  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: July 05, 2013, 09:42:16 PM
Cheers. I'll check it again tomorrow.


To be sure, my system lost connection three times. Once for a few seconds, the other for about 2 mins, and the other about 10-15 mins. Not sure if that would have cost me significantly, but I'm getting the impression not.

On long rounds like the 3+ hour ones disconnecting for a few minutes won't matter as long as it's at the beginning of the round.  Now if it's a short round or you disconnect near the end of the round then you lose out big time!  But... that's how Slush's anti-pool hopping rewards system works.
4  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: July 05, 2013, 09:35:11 PM
Hi, I'm mining for the first time, on Slush as it was frequently recommended. Still haven't got my head around the rewards systems. I've been running at 1.6Gh/s for about 10 hours, and the rewards are so far pretty low. I'm assuming this has to do with the scoring. How long should I wait until the reward rates are more representative?

I'm also running around 1.6Gh and here's what I've been getting:

18958   2013-07-05 15:50:22   3:10:51   56002768   4102   0.00180373   244940   25.33712441    60 confirmations left
18957   2013-07-05 12:39:31   0:04:38   1348242   103   0.00185563   244919   25.00910000    39 confirmations left
18956   2013-07-05 12:34:53   1:19:39   23519214   1678   0.00153073   244916   25.06413957    36 confirmations left
18955   2013-07-05 11:15:14   0:09:40   2823303   210   0.00190709   244905   25.08810000    25 confirmations left
18954   2013-07-05 11:05:34   1:27:29   25736694   1858   0.00181915   244903   25.28290685    23 confirmations left
18953   2013-07-05 09:38:05   3:03:50   53744650   4067   0.00187024   244896   25.21930000    16 confirmations left

Those 3 hr plus rounds aren't exactly helping anyone's rewards  Shocked

With the long rounds the PPS reward system might give you better results but overall I find that I earn more on Slush.  I'm too lazy to do the hard math right now, but in general Slush's reward system is ratio based (with aging to prevent pool hopping) so you can see that my results between round #18958 (0.00180 BTC) which was 3+ hours and #18957 (0.00185) which was 4+ minutes was about the same.  Now with PPS (on BTCGuild for example) you would get ~ 0.00444 BTC for round #18958 based on what their PPS rate is, but for #18957 you would have received 0.00011 which gives you about 0.00455 BTC versus Slush which gave 0.00365 BTC.  Not that great if you look at it.  However, if the rounds start getting shorter then you would start making more because a couple of short rounds like #18957 and  #18955 would earn you significantly more (e.g. 0.00185 vs. 0.00011 per round which is an order of magnitude more).  Since this is all still luck based, there is no set time to wait until rates are more representative.  I can only say wait and see.  Well, at least that's how I understand it.  I'm sure the more knowledgeable folks on here will do a better job explaining the results and correct me where I'm wrong.  There's probably a duration someone can calculate statistically with 95% certainty that you'll see more "representative results", but I'm guessing that's probably a large number.
5  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 24, 2013, 12:21:30 PM
It seems that payments from confirmations have stopped as well.  Damn.
6  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 24, 2013, 10:38:14 AM
Thanks for the replies, laserray, Humax.

Seems that neither leaving out the "http://" nor using stratum, nor any permutation of those solutions works.

I'm really confused because everything worked fine a couple days ago and the same exact command now seems to be failing.  I'm still able to join other pools too.

I have seen some "slush" victories on blockchain.info, so I don't think we're down.  At least, we haven't been down the whole time I've been suffering this problem.

Could something with my ISP have changed???
Well, funny that you mention it....
The stratum address should work except the pool is experiencing some issues right now (i.e. it went down for a lot of people not too long ago).  I'd wait until slush fixes things before giving the stratum address another go.  Also you can use direct IP if there's some name resolution issues going on.
7  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 24, 2013, 10:05:18 AM
Stratum error messages across the board over here.  We under attack again?

Same here... switched failover pool for now.
8  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 24, 2013, 09:24:29 AM
Maybe someone can help me with this.  I did my first mining two days ago with Slush.  I ran it for about a day.  Now, the same exact command line gives me the following error:

 [2013-06-24 11:33:05] Started cgminer 3.2.2
 [2013-06-24 11:33:05] Probing for an alive pool
 [2013-06-24 11:33:06] Pool 0 slow/down or URL or credentials invalid


./cgminer -o http://api.bitcoin.cz:8332 -u username.workername -p password -I 12 --temp-cutoff 80

What gives?


Why don't you use:

stratum.bitcoin.cz:3333

so your command line would be like the following:

./cgminer -o http://stratum.bitcoin.cz:3333 -u username.workername -p password -I 12 --temp-cutoff 80
9  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 21, 2013, 09:29:00 AM
.....Aaand this topic should be called [16000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested  Wink

the topic was changed to 12,000 so  am sure it will go to 16k sooner or later haha...

gotta love the post count anti OMG scammer count negater. 20+ posts and still 42. I will never get to 55 so I can collect BTC! Sad

16TH? It's now at 18.4TH... so crazy.
10  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 12, 2013, 02:52:18 PM
Total network hash rate has jumped by nearly 10% to over 154Thash/s since last night. Diff estimated to increase to over 18 million by Sunday.

Now this is interesting ! Asicminer activating a lot of new block erupters out of the factory ?
I wanted to invest in asic but as I said in another post I think that very soon 50gh/s will be the minimum to hash so I will refrain and sadly just watch very soon :/

I feel the same way.  With the pace at which the Hash rates are increasing, I just don't have enough funds to acquire the hardware to keep mining with a reasonable profit margin.  I'm guessing I'll have to shutdown buy the end of the year since I probably won't make enough to offset the cost of electricity and I'm running FPGAs.  I wonder how hard this is hurting the GPU miners. This is pretty depressing  Embarrassed and in the long run, such a huge jump might actually be detrimental to the bitcoin economy.  I know many people compare this to the switch from CPU mining to GPU mining, but that's somewhat different.  In the case of going from CPUs to GPUs, GPUs are readily available at any computer store.  ASICS on the other hand are hard to acquire.  Look at the questionable bidding wars over the erupter blades and USBs.  GPU prices never got that out of hand.  This huge jump could in fact alienate many newcomers and result in the dissolution of the bitcoin network.
11  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 11, 2013, 02:23:10 AM

It's combined.

1 Laptop (AMD A4 APU) - 25 MH
2 FPGAs (Lancelot) - 400 MH x 2 = 800 MH

I was surprised by my luck so I've used the last of my "hobby" funds to get 2 more FPGAs.  By the end of the month, I should get another 2 Lancelot boards which should put me at ~ 1.625 GH.  I don't think I'll ever get my J from BFL  Undecided so I'm not counting that into my future Hashing increase.


nice. where do you get the FPGA Lancelots from? they run standalone power supplies attached?
I presently am just throwing GPU's in dekstops that are idle around me at work.

I got the Lancelot FPGA(s) from BlackArrow @ www.cardreaderfactory.com

Each FPGA does ~ 400 MH and sells for $350 USD.  Unfortunately, they ship from Hong Kong so the shipping is a bit steep @ $70 USD.  Of course, you'd want to get the most out of your shipping so you'd get multiple items at once.  As for the power, they sell a power brick that can power 4 FPGA(s) for $20 USD.  Each FPGA is supposed to only use ~ 28 W which is pretty good compared to GPUs @ ~100W but not as good as the USB ASICs.  Now from a value comparison, it's about $1 per MH which is on par with the USB ASICs.  However, I'd argue that if Bitcoins ever disappears, you can always resale or repurpose FPGA(s).  You can't do that with ASICs.  Note that you can't do LTC with these FPGA(s) because LTC is memory intensive (that's more for GPUs).  Anyhow, I think a batch is set to go out by June 27th so you should hit them up fast if you want some.  I've had mine for about 2 weeks now and the quality is actually quite impressive (although they could improve the fans a bit).

Damn, BlackArrow should give me a referral bonus or something Smiley
12  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 11, 2013, 01:24:02 AM
Where can you find such stats?

The found blocks (the blocks you found) will show up on the "My account" page near the bottom.  Then you can go to the "Statistics" page to find out the round details for the given block.
13  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 11, 2013, 01:19:43 AM
nice another 4 in a row!

240789 240790 240791 240792

Hah... you can thank me for 240790  Grin

I'm surprised I found a block at only ~825MH and in about 4 min too.  Ain't luck a fickle mistress.

NICE! is that combined 825 or a single FPGA? gpu?

It's combined.

1 Laptop (AMD A4 APU) - 25 MH
2 FPGAs (Lancelot) - 400 MH x 2 = 800 MH

I was surprised by my luck so I've used the last of my "hobby" funds to get 2 more FPGAs.  By the end of the month, I should get another 2 Lancelot boards which should put me at ~ 1.625 GH.  I don't think I'll ever get my J from BFL  Undecided so I'm not counting that into my future Hashing increase.
14  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 11, 2013, 12:48:16 AM
nice another 4 in a row!

240789 240790 240791 240792

Hah... you can thank me for 240790  Grin

I'm surprised I found a block at only ~825MH and in about 4 min too.  Ain't luck a fickle mistress.
15  Bitcoin / Pools / Re: [12000 GH/s] Slush's Pool (mining.bitcoin.cz); TX FEES + UserDiff; ASIC tested on: June 09, 2013, 08:19:15 AM
Anyone else getting some funny results with 18481 as well?

18481   2013-06-09 07:27:23   1:55:53   21128042   1295   0.00000289   0.00001035   240547   25.03220000    98 confirmations left

for comparison

18467   2013-06-08 07:46:53   1:51:32   22029843   1276   0.00133798   none   240367   25.11510000    confirmed
18455   2013-06-07 15:19:28   1:54:51   22203035   1386   0.00160477   none   240244   25.15792127    confirmed


I don't think I lost connectivity but I can't say for sure since I don't log/monitor 24/7.  It seems like the pool has been acting kinda weird with the scoring.  I wonder what's up.
16  Other / Beginners & Help / Re: GPUs vs Block Erupter USBs? on: June 06, 2013, 05:08:43 AM

i think imma just buy 1 of the usb things and a rasberry pi, leave it and forget it. lol check back once a month and order pizza with the winnings Tongue

Thats a good way to look at it ...but even then you might not have enough for a pizza lol

Here's some more food for thought...

Assuming the difficulty doesn't increase too much (it's already climbing pretty fast) and that you're part of a fairly lucky pool (a lot of ifs), at ~300 MH you should see something like 0.01 BTC a day.

That fastest way to get pizza is to use a service that takes BTC (and charges a fee) to buy pizza because you'll have to wait a long time to hit minimum amounts to do money conversion at most exchanges.  Now let's say the pizza costs $10 and the fee is $2 @ a rate of $100 per BTC (for simplicity), you'll need 0.12 BTC.  That's 12 days at the previous rate to be able to get pizza, so 2 pizzas a month which is not bad.

However, considering that you spent 2 BTC on the hardware you could just buy yourself 16 pizzas to begin with.  That's 2 pizzas for the next 8 months.
17  Other / Beginners & Help / Re: GPUs vs Block Erupter USBs? on: June 06, 2013, 04:27:09 AM
Granted the question is GPU vs. Eruptor USB, but has anyone considered how fpga(s) stack against the Eruptor USB.  I'm actually using fpga(s) because:

1) They're similar to the Eruptor USB at ~ $1 per MH
2) They use more electricity than the Eruptor USB, but not that much compared to GPUs  (my Lancelots use 28W compared with GPUs at 100W)
3) And this is the key point for me... fpga(s) can be repurposed (assuming you know how to edit bitstreams) and have more resale value compared to the Eruptor USB.

Unless the Eruptor USB comes down more in price... it's not worth it from my viewpoint because it has no resale value and the amount of time to recover costs is not significantly better than fpga(s).

Anyhow my 2 bitcents...
18  Other / Beginners & Help / Re: What's your Mhash/s? (Pissing contest here) on: June 06, 2013, 03:54:13 AM
It seems most people here are running GPU rigs.  That's cool if you use your GPUs to game and mine as hobby, but the profit margin is slimmer due to the high power consumption of the GPUs (unless of course, you have free electricity - solar, wind, etc.)

My hashing comes mostly from fpga(s)

1 AMD A4 Laptop - AMD Radeon 6480G - set at 25 MH @ ~ 65 W
2 Lancelot FPGAs boards - 2 Spartan 6(s) per board - 400 MH x 2 @ ~56 W combined

currently hashing at 825 MH

I've got 2 more Lancelot boards planned for the end of the month to bring me up to 1625 MH

I purchased an ASIC from BFL a month or two ago which I'll probably never see, but if by some miracle it gets to me it'll be 5 GH so.... 6625 MH
19  Other / Beginners & Help / Re: why can't i post outside this section? I have 5 posts.... on: June 06, 2013, 12:43:26 AM
Did you also have 4 hours of online time in addition to the 5 posts?
20  Other / Beginners & Help / Re: Noob level up? on: June 06, 2013, 12:40:36 AM
So does the 4 hour ticker start after the 5th post or is it concurrent with account creation?  I appologize if this has already been answered.
Pages: [1] 2 »
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!