Topic: [POOL][Scrypt][Scrypt-N][X11] Profit switching pool - wafflepool.com - page 58. (Read 465716 times)

PoolWaffle
why Entropycoin was disabled?

WafflePool has consistently been the largest DRK coin pool for over a day now!

http://drk.poolhash.org/poolhash.html

(WafflePool is pool_unknown_80).

Read page 234

any1 knows a good cpu x11 miner?

There is a FAQ: http://wafflepool.com/faq#normalized

 

Cost of electricity and concern over heat vary between mining operations.  Some people won't care about power or heat (free or cheap power, and large areas or good cooling), and some people will care a lot about power/heat (high power cost, or multiple rigs in condos/apartments).

I think for the summer you should do an X11/X13 auto switching pool, because about the same power, low heat, low electric, and easy to switch.
Then for the winter do the Scrypt/Scrypt-N/Keccak auto switching pool which all run more electric bill and more heat (but don't care in winter)

No need for user registration.  Better to have people sign a generated message from their BTC address, I think.  (e.g. how sites like SimpleVert take Merge-Mined addresses http://www.simplevert.com/set_merge/VkbHY8ua2TjxdL7gY2uMfCz3TxMzMPgmRR/MON or allow fees to be increased http://www.simplevert.com/set_donation/VkbHY8ua2TjxdL7gY2uMfCz3TxMzMPgmRR).

Hold-back (no auto-trade) coins could be implemented along with user-weighted algorithm preferences via this method, and still no user-signup for those that want easy quick-start mining.

İ have very high rejected percentage. 77.61 MH/s (24.5%)in stratum+tcp://eu.wafflepool.com:3331 .Otherpools max i had %1 in same settings.What's wrong ??

i have some 270s mining x11 just fine getting about 1.9Mh/s per card, getting shares.  

my 290x is running at 4.11Mh/s, but i am getting no accepted shares.
here are my settings:

"intensity" : "15",
"worksize" : "128",
"kernel" : "x11mod",
"lookup-gap" : "0",
"thread-concurrency" : "24550",
"shaders" : "0",
"gpu-threads" : "1"

all i keep seeing is diff set to xx and stratum detected new block

help?

edit: one thing i should mention is that the pc with 290x has the latest AMD drivers installed, but it seems like thats a good thing

Please make sure you have upgraded the Entropycoin Wallet to v0.9.2. THe is a hard fork in less than 200 blocks. I sent an email a few days ago and didn't receive a response. Just want to make sure you're aware.

http://entropycoins.com/

Thank You!

don't you think it is time to move to gpuwafflepool with registration and individual preference settings for upcoming smart algo-switching?
and to keep wafflepool scrypt-asic friendly without registration.

Right, its definitely a hack, just abusing the way that failover works.  I'm sure it will get sorted out, but its a great answer for the short-term

Yup, I'm sure you would find a way to set it up without extra duplication on your end, but I meant cgminer config - I need to set up a whole "pool" section for each algo, when the difference between all these "pools" is basically one digit in the port number. Then copy pasta a few times for US east, US west, and any other backup pool. It's verbose to say the least and consequently prone to errors and pita to maintain. I don't think sgminer devs had multipools in mind when they came up with this.

I don't think it would require an extra pool for each algorithm, just proxied ports that are enabled/disabled depending on profitability.  But I agree, its a good bit of a hack (first iterations of things always are)


It would 100% be opt-in to use (different set of ports), a lot of our miners physically can't use multi-algo, considering a good number use Scrypt ASICs

If you do end up implementing an algo switcher, I would like to be able to opt-out of it. I am mining X11 right now due to power constraints, if I switch back to scrypt I'll pop the breaker.

I wouldn't call it that, it's a hack really, hopefully temporary until a better solution evolves. It doesn't allow to pick which algorithms to use (along with the complete lack of user preferences as you mentioned), it only works with AMD GPUs, it requires to set up a separate "pool" for each algorithm (times 2 or 3 with backups), and is generally not scalable. It's not hard to imagine that a possible future path for GPU mining might be a growing/changing list of algos to choose from, so there has to be a better way to do the choosing.

@PW
Please consider algorithm groestl also

Good question

X13 is the obvious first step.  I'm in the zone with adding the new algorithms, and X13 shouldn't be much work to add from X11.  I'm out of town this weekend, so it probably won't be finished until next week, but should be pretty quick.

After that, like you said, combining payouts has been a major request, and shouldn't be much work.  It should have been done by now, but I like to be extra cautious on things like payouts.  Don't want anything to screw up there

Past that, I like the idea of the algo-switching, but it does require some extra thinking.  The way it is done with NiceHash is really neat (props to them for the idea, and funding the custom version of sgminer), but at this point, the new version of sgminer seems to have some issues.  Ideally by the time I've got the other stuff done, the major kinks will be worked out in the new sgminer version, and we can look into an algo-switching version of WP.  The actual algo-switching part isn't too tough (pretty easy in fact), the bigger issue is figuring out how to do the switching based on user preferences.  For example, if your miner can do 1MHs Scrypt, and 6MHs X11 (normal is 4MHs X11 for a 1MH scrypt), you would need to be treated differently in terms of what is profitable than someone with 1MH/4MH.  There is also the difference of preference of power consumption.  Some people would like to bias their miners towards X11/X13 because of the power savings, some users don't care, and would prefer to mine whatever is the most profitable (regardless of power costs).

Doing algo-switching on a generic level (preset values) is reasonably easy.  Doing algo-switching with user preferences becomes quite a bit more difficult