[POOL][Scrypt][Scrypt-N][X11] Profit switching pool - wafflepool.com

[utahjohn]

We have another newcomer to wafflepool:

im proud to announce the open beta of
http://multipool.bit.diamonds/
the ticket/lottery system will follow when open beta ends
but beside that its full functional u will earn diamonds already
finally a way earn diamonds with ur sha256 and scrypt asics too

[utahjohn]

@PW
Looks like u might want to bump up vardiff on the list of things to get working ...

[Ignition75]

Problems with stratum?  My charts are up and down like a yo yo...

[phzi]

@PW
Looks like u might want to bump up vardiff on the list of things to get working ...

Oh, please no.  Vardiff is just annoying, IMO.  At least, I've never seen a vardiff implementation that stabilizes on a difficulty well enough - they all hop around too much.

[utahjohn]

Problems with stratum?  My charts are up and down like a yo yo...

Looks like normal variance to me ...

[goodluckpool]

Problems with stratum?  My charts are up and down like a yo yo...

Hi there, I thought you are going with p2pool

Hint, set a lower difficulty in password like d=0.05, if you use 280x, so your hash rate will be more steady.
But don't set it too low.

My testing shown difficulty should be set according to the hash rate of per card, not the whole rig, for example, if you have 4 280x, setting a 0.16 difficulty will waste tons of your hash rate because waffle pool is fast at finding blocks, if it takes over 1 or 2 minute for each a card to submit a work, there is a big chance a new block has been found and your wasted lots of GPU time hashing the old block.  So set d=0.05 no matter how many 280x you have in your rig will work better.

I am talking about X11. The theory is each card was actually working individually with 2 threads (-g 2) on solving the assigned work, they don't work together, 4 cards in a rig means you have 8 threads working independently, but the difficulty set in stratum protocol was assigned to each card, not divided among the number of cards.

so, for example, rig 1 has 1 280x, I use d=0.05, rig 2 has 6 280x, I still use d=0.05, each card (thread) will submit a hash in same amount of time.

but if you use d=0.30 on that 6 280x rig, each card will take 2-3 minutes to submit a hash and there is a good chance it will be stale or being interrupted in the middle of working because a new block has been found.

correct me if I am wrong.  

[Ignition75]

Problems with stratum?  My charts are up and down like a yo yo...

Hi there, I thought you are going with p2pool

Hint, set a lower difficulty in password like d=0.05, if you use 280x, so your hash rate will be more steady.
But don't set it too low.

My testing shown difficulty should be set according to the hash rate of per card, not the whole rig, for example, if you have 4 280x, setting a 0.16 difficulty will waste tons of your hash rate because waffle pool is fast at finding blocks, if it takes over 1 or 2 minute for each a card to submit a work, there is a big chance a new block has been found and your wasted lots of GPU time hashing the old block.  So set d=0.05 no matter how many 280x you have in your rig will work better.

I am talking about X11. The theory is each card was actually working individually with 2 threads (-g 2) on solving the assigned work, they don't work together, 4 cards in a rig means you have 8 threads working independently, but the difficulty set in stratum protocol was assigned to each card, not divided among the number of cards.

so, for example, rig 1 has 1 280x, I use d=0.05, rig 2 has 6 280x, I still use d=0.05, each card (thread) will submit a hash in same amount of time.

but if you use d=0.30 on that 6 280x rig, each card will take 2-3 minutes to submit a hash and there is a good chance it will be stale or being interrupted in the middle of working because a new block has been found.

correct me if I am wrong.  

I'm running tests at the moment, multipools vs P2Pool vs pool mining DRK directly...  Whatever gets me more DRK is what I will settle on.

Graphs were smooth as silk for the first 18 hours, woke up this morning to find massive variances.  I've currently got diff on 0.16, each rig is around 16 Mh/s on x11.

[goodluckpool]

Problems with stratum?  My charts are up and down like a yo yo...

Hi there, I thought you are going with p2pool

Hint, set a lower difficulty in password like d=0.05, if you use 280x, so your hash rate will be more steady.
But don't set it too low.

My testing shown difficulty should be set according to the hash rate of per card, not the whole rig, for example, if you have 4 280x, setting a 0.16 difficulty will waste tons of your hash rate because waffle pool is fast at finding blocks, if it takes over 1 or 2 minute for each a card to submit a work, there is a big chance a new block has been found and your wasted lots of GPU time hashing the old block.  So set d=0.05 no matter how many 280x you have in your rig will work better.

I am talking about X11. The theory is each card was actually working individually with 2 threads (-g 2) on solving the assigned work, they don't work together, 4 cards in a rig means you have 8 threads working independently, but the difficulty set in stratum protocol was assigned to each card, not divided among the number of cards.

so, for example, rig 1 has 1 280x, I use d=0.05, rig 2 has 6 280x, I still use d=0.05, each card (thread) will submit a hash in same amount of time.

but if you use d=0.30 on that 6 280x rig, each card will take 2-3 minutes to submit a hash and there is a good chance it will be stale or being interrupted in the middle of working because a new block has been found.

correct me if I am wrong.  

I'm running tests at the moment, multipools vs P2Pool vs pool mining DRK directly...  Whatever gets me more DRK is what I will settle on.

Graphs were smooth as silk for the first 18 hours, woke up this morning to find massive variances.  I've currently got diff on 0.16, each rig is around 16 Mh/s on x11.

could be
1) network issue,
2) first 18 hour he was mining high difficulty coins, now he is mining low difficulty fast new coins, that will cause the massive variance issue with d=0.16 , consider that.
3) check the scale of the graph

[Ignition75]

could be
1) network issue,
2) first 18 hour he was mining high difficulty coins, now he is mining low difficulty fast new coins, that will cause the massive variance issue with d=0.16 , consider that.
3) check the scale of the graph

Yes I thought about the high diff coin reason.  It's also happened on the x13 port which is why I thought it may have been something else.  I've got half my rigs on x11 and half on x13 (which is also another test of mine), it's pretty even so far.

I just took my x11 rigs down to a diff of .08, we'll see if that makes any difference.

[goodluckpool]

could be
1) network issue,
2) first 18 hour he was mining high difficulty coins, now he is mining low difficulty fast new coins, that will cause the massive variance issue with d=0.16 , consider that.
3) check the scale of the graph

Yes I thought about the high diff coin reason.  It's also happened on the x13 port which is why I thought it may have been something else.  I've got half my rigs on x11 and half on x13 (which is also another test of mine), it's pretty even so far.

I just took my x11 rigs down to a diff of .08, we'll see if that makes any difference.

my test shown X13 is not that good because although in old days:

"In the case of algorithms besides Scrypt, we multiply by the average change in speed when a miner switches to that algorithm. Scrypt-N is 0.47 (a 1MHs miner on Scrypt does about 470kHs on Scrypt-N). X11 is 4.0 (a 1MHs Scrypt miner will normally do around 4MHs on X11). X13 is 3.5. This lets you look quickly at the "vsLTC" column to see how well each algorithm is performing on a normalized scale. "

The new situation is using best miner today, a 280X can do 4.2MHs on X11 but only 2.9-3.0 on X13. So I stick with X11 for now, and some new X13 coins became worthless after mining.
(except today, the X11 number looks bad)

[zSprawl]

We have another newcomer to wafflepool:

im proud to announce the open beta of
http://multipool.bit.diamonds/
the ticket/lottery system will follow when open beta ends
but beside that its full functional u will earn diamonds already
finally a way earn diamonds with ur sha256 and scrypt asics too

lol, geez.

So the question is, do you block people who do this, or allow it?

[Ignition75]

my test shown X13 is not that good because although in old days:

"In the case of algorithms besides Scrypt, we multiply by the average change in speed when a miner switches to that algorithm. Scrypt-N is 0.47 (a 1MHs miner on Scrypt does about 470kHs on Scrypt-N). X11 is 4.0 (a 1MHs Scrypt miner will normally do around 4MHs on X11). X13 is 3.5. This lets you look quickly at the "vsLTC" column to see how well each algorithm is performing on a normalized scale. "

The new situation is using best miner today, a 280X can do 4.2MHs on X11 but only 2.9-3.0 on X13. So I stick with X11 for now, and some new X13 coins became worthless after mining.
(except today, the X11 number looks bad)

It's interesting you say that, I got 15 rigs on each algo, x11 = 250 Mhs, x13 = 200 Mhs...

Total payout so far (I've been on for roughly 24 hours) - x11 = 0.09007237 & x13 = 0.10509369...

Need more time to come up with meaningful stats, I'm going to give it a week...

[Ignition75]

I gotta take my hat off to whomever runs this pool... Lowest number of rejects from any multipool I've ever tried...

I'm averaging about 0.3% rejects from each rig... That's pretty damn impressive...

Go somewhere like clevermining and you'll be lucky to sit on 5%...

[goodluckpool]

I gotta take my hat off to whomever runs this pool... Lowest number of rejects from any multipool I've ever tried...

I'm averaging about 0.3% rejects from each rig... That's pretty damn impressive...

Go somewhere like clevermining and you'll be lucky to sit on 5%...

told u to try this one, and use the proceed to buy back DRK

this pool can take on extra 80G of chinese X11 hash with zero sweat.

[goodluckpool]

my test shown X13 is not that good because although in old days:

"In the case of algorithms besides Scrypt, we multiply by the average change in speed when a miner switches to that algorithm. Scrypt-N is 0.47 (a 1MHs miner on Scrypt does about 470kHs on Scrypt-N). X11 is 4.0 (a 1MHs Scrypt miner will normally do around 4MHs on X11). X13 is 3.5. This lets you look quickly at the "vsLTC" column to see how well each algorithm is performing on a normalized scale. "

The new situation is using best miner today, a 280X can do 4.2MHs on X11 but only 2.9-3.0 on X13. So I stick with X11 for now, and some new X13 coins became worthless after mining.
(except today, the X11 number looks bad)

It's interesting you say that, I got 15 rigs on each algo, x11 = 250 Mhs, x13 = 200 Mhs...

Total payout so far (I've been on for roughly 24 hours) - x11 = 0.09007237 & x13 = 0.10509369...

Need more time to come up with meaningful stats, I'm going to give it a week...

are those payout earned payout or including unconverted payout?

[Ignition75]

my test shown X13 is not that good because although in old days:

"In the case of algorithms besides Scrypt, we multiply by the average change in speed when a miner switches to that algorithm. Scrypt-N is 0.47 (a 1MHs miner on Scrypt does about 470kHs on Scrypt-N). X11 is 4.0 (a 1MHs Scrypt miner will normally do around 4MHs on X11). X13 is 3.5. This lets you look quickly at the "vsLTC" column to see how well each algorithm is performing on a normalized scale. "

The new situation is using best miner today, a 280X can do 4.2MHs on X11 but only 2.9-3.0 on X13. So I stick with X11 for now, and some new X13 coins became worthless after mining.
(except today, the X11 number looks bad)

It's interesting you say that, I got 15 rigs on each algo, x11 = 250 Mhs, x13 = 200 Mhs...

Total payout so far (I've been on for roughly 24 hours) - x11 = 0.09007237 & x13 = 0.10509369...

Need more time to come up with meaningful stats, I'm going to give it a week...

are those payout earned payout or including unconverted payout?

It's including unconverted, just taking the total figures at this stage.

I tell a lie as well, x13 is hovering around 190 Mh/s, I haven't spent a lot of time tweaking but you can normally get x13 to 80% of x11...

[Ignition75]

It's interesting you say that, I got 15 rigs on each algo, x11 = 250 Mhs, x13 = 200 Mhs...

Total payout so far (I've been on for roughly 24 hours) - x11 = 0.09007237 & x13 = 0.10509369...

Need more time to come up with meaningful stats, I'm going to give it a week...

are those payout earned payout or including unconverted payout?

I evened out the hashrates, rigs 1-15 have a couple of 6 x 280x rigs and some 4 x 290 rigs, where rigs 16-30 are purely 6 x 270x.

So I got similar comparable hashrates between x11 and x13 now, 14 rigs on x11 and 16 rigs on x13, I will run this test for 1 week and report my findings.

Hopefully buying a lot of DRK along the way 

[goodluckpool]

Mr. poolwaffle:

do you have CYC | Conspiracycoin in the X11 choice? it seems quite profitable at this low difficulty according to stats from other multipools, and with our 20G we can slice through it like knife in butter.

[toxic0n]

New Waffle Widget update

- Added support for X11/X13
- You can now view the total hash rate per algorithm or specify individual workers by name for each widget.

Enjoy!

[duboisi]

Mr. poolwaffle:

do you have CYC | Conspiracycoin in the X11 choice? it seems quite profitable at this low difficulty according to stats from other multipools, and with our 20G we can slice through it like knife in butter.

CYC has less than 8Gh/s, swinging 20G over would increase the diff 4x and profitability down to 25%.
WP is too big for recent high profitable coins with low network hash.