Starting a Lan pool mining compared to using bitcoind

[kindle]

Hi I am curious, does running a local pool server distributing out 'shares' more efficient compared to using the standard bitcoind? I noticed the increase shares generation per gpu worker in a mining pool compared to solo mining using bitcoind. I was wondering if this idea of breaking the work into smaller shares like what mining pools are doing could be applied to solo mining if let say I have 6 gpu and more running. Would using normal bitcoind in solo mining cause the same work to be processed redundantly? For example, I recalled slush mentioned the need to register each gpu into individual workers on the website.

Could using remote mining server tool like that developed by puddinpop be the way to go?

[JorgePasada]

I'd also be interested in knowing more about this. I'm thinking about starting my own 'local pool'

[xf2_org]

There is already an open source pool backend:  https://github.com/jgarzik/pushpool

And (see signature) xf2.org offers private mining pools as well.

[JorgePasada]

There is already an open source pool backend:  https://github.com/jgarzik/pushpool

And (see signature) xf2.org offers private mining pools as well.

Thanks for this info. So question (to make sure I'm understanding this properly)

Say I have 10 cards each operating at 400khash for a total of 4ghash.

Am I correct in assuming that if I just have each card mining on it's own (IE 10 x 400khash) there's a chance that one or more of those 10 may be trying the same combination of hashes and thus reducing my overall effectiveness?

Corollary, is it then correct to assume the reason people pool their own cards it because it's more akin to having 1x 4ghash system working on the problem with no chance of duplicates, whereas the other way you might be wasting GPU power theoretically?

Does that make sense to anyone or do I not understand this at all.

[xf2_org]

Am I correct in assuming that if I just have each card mining on it's own (IE 10 x 400khash) there's a chance that one or more of those 10 may be trying the same combination of hashes and thus reducing my overall effectiveness?

No, that's not correct.  The work you receive is always unique.

You can mine directly against bitcoind, or put a proxy in front of bitcoind (pool server) for better stats and management.

[kindle]

No, that's not correct.  The work you receive is always unique.

You can mine directly against bitcoind, or put a proxy in front of bitcoind (pool server) for better stats and management.

Hi does the proxy in this case a pool mining server provide better allocation of work to the gpus compared to bitcoind interms of probability of solving a block. For an example, the work is divided into shares before sending to the GPU miners to hash, rather than the distributing the same entire workload to the number of GPUs available thereby creating a redundancy? (correct me if I am wrong as that is the perception I have now)

[JorgePasada]

Am I correct in assuming that if I just have each card mining on it's own (IE 10 x 400khash) there's a chance that one or more of those 10 may be trying the same combination of hashes and thus reducing my overall effectiveness?

No, that's not correct.  The work you receive is always unique.

You can mine directly against bitcoind, or put a proxy in front of bitcoind (pool server) for better stats and management.

So what's up with things I've heard about people working on 'dead blocks' or blocks that have already been solved?

Sorry for the dumb questions, just trying to wrap my head around this.

[xf2_org]

No, that's not correct.  The work you receive is always unique.

You can mine directly against bitcoind, or put a proxy in front of bitcoind (pool server) for better stats and management.

Hi does the proxy in this case a pool mining server provide better allocation of work to the gpus compared to bitcoind interms of probability of solving a block.

The probability of solving a block is the same for everybody, and you cannot change that.  It's a function of being a cryptographic hash.

For an example, the work is divided into shares before sending to the GPU miners to hash, rather than the distributing the same entire workload to the number of GPUs available thereby creating a redundancy? (correct me if I am wrong as that is the perception I have now)

There is never any redundant work performed.

[xf2_org]

So what's up with things I've heard about people working on 'dead blocks' or blocks that have already been solved?

At any second, a block may be "solved."  This means that everyone else in the world working on that block must stop, and restart their work.  Continuing to work after that point is known as working on a "stale block" because it is old data, and old transactions.