Slush vs Deepbit

[gigabytecoin]

Just got home, and noticed that one of the computers was running about 100MH/s slower than the other for some reason...

Looks like since Friday at about 8PM (both started at the same time), I received 18.19 BTC from slush's pool and 21.46 from deepbit... Given that I didn't have a way to check the MH/s average on slush's pool, and that when I got home, it was running a bit slower than the other machine, I have to say my results are inconclusive...

-EP

I would have to agree.

Which computer was running 100mh/s slower? deepbit or slush's?

[rezin777]

Just got home, and noticed that one of the computers was running about 100MH/s slower than the other for some reason...

Looks like since Friday at about 8PM (both started at the same time), I received 18.19 BTC from slush's pool and 21.46 from deepbit... Given that I didn't have a way to check the MH/s average on slush's pool, and that when I got home, it was running a bit slower than the other machine, I have to say my results are inconclusive...

-EP

I would have to agree.

Which computer was running 100mh/s slower? deepbit or slush's?

Given that I didn't have a way to check the MH/s average on slush's pool, and that when I got home, it was running a bit slower than the other machine

[Fiyasko]

Just because you cant check an average doesnt mean you cant tell how fast thier going, Wich miner was 100mhash slower when you got home?, We are all assuming it was the Slush miner that slowed down

[[Tycho]]

Just because you cant check an average doesnt mean you cant tell how fast thier going, Wich miner was 100mhash slower when you got home?, We are all assuming it was the Slush miner that slowed down

As i understand, he doesn't knows for how long it was slower.

[nster]

1) The message about cooling was an obvious joke,
2) One of slush's bitcoinds was not working all the night which caused 502 error on site, so no lies here (this didn't affected mining).

I'm using both your pool and Slush's at this moment, both are ok.
Anyway, opposite to Slush, you are always bashing your competition.
Not good for your own reputation, I think. 

what are you talking about >.> As far as I have seen, he does not bash the competition at all. and he is right, Slush's WEBSITE was down. Mining was fine. WEBSITE WEBSITE WEBSITE NOT MINING.... WAS WAS WAS = PAST TENSE

He was not bashing the competition, just stating the facts, and Slush himself would not deny this, I bet it was even talked about it in his thread.

[gusti]

I'm also comparing both pools, running a 5970 against each it for the last 3 days.
Swaping cards once a day, to lower any divergence error.
At this moment, Slush's is giving me 20% more compared to Deepbit, on the total reward.

[zoro]

could you please perform a comparison between pps and slush?
i find it logical for a larger pool to give more if it has more possibilities to find blocks than a smaller pool.
but pps is totally independent from block finding

[sc8nt4u]

I'm also comparing both pools, running a 5970 against each it for the last 3 days.
Swaping cards once a day, to lower any divergence error.
At this moment, Slush's is giving me 20% more compared to Deepbit, on the total reward.

I have a 5970 per pool too and it's been the case so far, but I'm only through about 20k shares for each pool.

[gusti]

could you please perform a comparison between pps and slush?
i find it logical for a larger pool to give more if it has more possibilities to find blocks than a smaller pool.
but pps is totally independent from block finding

just switched to pps on deepbit
until this moment, Slush's is 23% up
let's see if numbers run closer from now on.

[gusti]

could you please perform a comparison between pps and slush?
i find it logical for a larger pool to give more if it has more possibilities to find blocks than a smaller pool.
but pps is totally independent from block finding

just switched to pps on deepbit
until this moment, Slush's is 23% up
let's see if numbers run closer from now on.

Sorry, my mistake, I was taking unconfirmed figures on Slush's.
I can say now that figures on both pools are very similar, maybe 2% difference.
I will extend the test to confirm that.

[nster]

could you please perform a comparison between pps and slush?
i find it logical for a larger pool to give more if it has more possibilities to find blocks than a smaller pool.
but pps is totally independent from block finding

Your logic is not logical. in smaller pools, you get a bigger share of the block, but you find blocks less frequently, compared to larger pools. In the end, the size only affects the consistency of the payouts.

PPS is not something people with GPUs should use IMO. 10% fee will take a fair share of your BTC. The proportional of deepbit is nice though, as it is 3% fee, which is reasonable considering it is the 2nd largest pool and the one with most features.

[EPiSKiNG]

Just got home, and noticed that one of the computers was running about 100MH/s slower than the other for some reason...

Looks like since Friday at about 8PM (both started at the same time), I received 18.19 BTC from slush's pool and 21.46 from deepbit... Given that I didn't have a way to check the MH/s average on slush's pool, and that when I got home, it was running a bit slower than the other machine, I have to say my results are inconclusive...

-EP

I would have to agree.

Which computer was running 100mh/s slower? deepbit or slush's?

Upon more analysis, slush's was actually about 40mh/s slower than the box I had on deepbit's (I think it had to do with the AMD platform vs Intel)

-EP

[zoro]

i had a comparison between pps and proportional in deepbit. pps gave me less!
but i think this is a matter of luck. if many blocks are found by the pool, pps is out of the question!
the good think about deepbit is the payout also for invalid blocks! i wonder if any losses can occur to this pool by this method!

[[Tycho]]

i had a comparison between pps and proportional in deepbit. pps gave me less!

That's pretty expected since PPS price equals to 10% fee and proportional's fee is 3%

the good think about deepbit is the payout also for invalid blocks! i wonder if any losses can occur to this pool by this method!

You can check it at pool statistics history if you are talking about some other pool.

Thanks :)

[sc8nt4u]

Well I'm at around 30k shares per pool and here are just a few thoughts so far.

My average ping to Tycho's pool ranged between 100-110ms.
51 Stale Shares
32171 Shares Submitted
0.159% of my shares are stale shares

My average ping to Slush's pool ranged between 140-160ms.
199 Stale Shares
31756 Shares Submitted
0.626% of my shares are stale shares

Last 1000 blocks up to block 2872 on Slush's pool had 13 invalids or 1.3%.

Slush's Pool:

For every 50 BTC you make, 1.3% is lost to invalid shares and no one gets BTC
- (50.00 BTC * .013) = 0.65 BTC lost

From that 50 BTC, 2% is given to Slush
- (50.00 BTC * .02) =  1.00 BTC lost

From that 50 BTC, you will need to submit 0.626% more shares to account for the stale shares
- (50.00 BTC * .00626) =  0.313 BTC lost

Total ~ 48.04 BTC expected

Tycho's Pool

For every 50 BTC you make, 3% is lost to Tycho's pool fee
- 50.00 BTC * .03 = 1.50 BTC lost

From that 50 BTC, you will need to submit 0.159% more shares to account for the stale shares
- 50.00 BTC * 0.00159 = 0.0795 BTC lost

Total ~ 48.42 BTC expected



TL;DR Slush's pool is expected to pay out 48.04 BTC per 50 BTC generated, while Tycho's pool is expected to pay out 48.42 BTC. The largest determining factor in this discrepancy is that invalid blocks are paid out in Tycho's pool and accounts for a 1.3% loss in BTC in the last 1000 blocks from Slush's pool [Blocks 1872-2872].

[Fiyasko]

Are you willing to give an estimate of the total value ß gained from each? (you can use fake numbers if the ratios stay the same)

[[Tycho]]

Are you willing to give an estimate of the total value ß gained from each? (you can use fake numbers if the ratios stay the same)

You can compare average number of shares per block in both pools for a long enough timespan.
Comparing reward per short period can't be precise enough since we had some days with luck ~30% more or less than average, so this can give more than 50% error.

But, of course, you can give your numbers.

[slush]

Last 1000 blocks up to block 2872 on Slush's pool had 13 invalids or 1.3%.

It's 12 invalids, the #116884 was temporarily marked as invalid because I had to check its generation by self. Now it is valid and reward is distributed between workers.

The largest determining factor in this discrepancy is that invalid blocks

I improved server infrastructure to process block distribution faster. Looks like it helped, last 150 blocks (where I made changes) are without single invalid block. Of course I didn't solved it 100%, some invalid blocks are normal, but the ratio should be much better.

[gusti]

any news yet on the comparitive perfromance of the pools?

Finished a 5 days testing on both, no negligible differences.

[Fiyasko]

Thankyou Gusti!