Quote from: flipme on April 04, 2014, 06:10:50 PM

Quote from: cryptohunter on April 04, 2014, 05:46:56 PM

Quote from: flipme on April 04, 2014, 05:39:55 PM

Quote from: cryptohunter on April 04, 2014, 05:28:27 PM

Quote from: iopq on April 04, 2014, 05:04:12 PM

Quote from: cryptohunter on April 04, 2014, 04:54:32 PM

Quote from: iopq on April 04, 2014, 04:41:35 PM

Quote from: cryptohunter on April 04, 2014, 03:29:30 PM

Quote from: softron on April 04, 2014, 03:13:58 PM

It quite good algo less energy consumption. Less noise try mining hirocoin

the algo = less energy?? you are still posting this after reading the entire thread?

yes

compare SHA-256 mining with GPU to scrypt
SHA-256 was optimized to the bone with Bitcoin, it was over 99% efficient compared to a "perfect" miner that didn't waste any ops anywhere because it only had a few ops here and there that were necessary but not part of the algorithm

yet it still takes less energy than mining scrypt, when scrypt possibly isn't even as optimized (there aren't several competing scrypt kernels afaik, but for Bitcoin there was fatk, poclbm, diablo, etc.)

Ok so that sounds more like the replies we are after. Let me just understand this.

sha256 miners although almost 100% efficient running on gpu - draw less electricity and creates less heat than scrypt... can i ask by exactly how much less electricity? i've never mined sha256...are we talking 50%

sha256 mining you are sure was optimised to almost 100%? what level of optimisation do you consider x11 miners to have at this point?

Can this therefore be a direct indication and something we can even apply to x11 ?

Also is there any reason at all why x11 could or should me more efficient than QRK? in terms of heat and electricity used even when the cards calculations resources are maxed and fully optimised?

So far the answer i am getting on that is NO. However if you really think it is possible that is something we need to investigate and consider now.

the very first SHA-256 miner was only maybe 5% less efficient than the best one (poclbm was giving me good hash rates before a lot of these newer kernels optimized it by a few more percent)
people just found 1% improvements here and there a few times and after a while there were no optimizations left
some optimizations reduced the total instruction count by one out of 1300 (less than 0.1% improvement)

SHA-256 takes like 20% less energy to mine off the top of my head because it doesn't use the GPU ram as much
so I'm sure there are algorithms that take even less GPU power because they rape one part of the chip (bottleneck) and don't touch the rest of it much

OK so it seems that some algos will bottleneck certain parts of the gpu before others so since other parts are left redundant it draws less energy. I did actually ask that exact thing earlier in the thread but nobody took up on it. Okay so let's be fair... so algos on x11 could be running as fast as they ever will on our gpus and will burn less energy than scrypt? is that a fair assumption?

However is 50% sounding correct? that does sound like some miner tweaks and mods could be very useful..the bottlenecks seem rather large here?

Also since none of these algos in this mix or qrks mix are memory hard, it seems those 2 should be equally as efficient since the algos will be creating these same bottlenecks and stoping the memory getting over taxed? or very very close too close to seperate really. So x11 is not ground breaking with regard the efficiency aspect since we can see sha256 and all the other algos in QRK and x11 are possibly just as easy on our electricity usage.

So the super new efficiency story seems to be a good marketing at best or perhaps just wishful thinking.... but perhaps not as open to super tweaking of the miner as we thought could be taking place.

So we rule out going x11 based on efficiency alone...

Let's do it like this.... a kind of battle of the algos....

so for efficiency running on gpu with current miners - - any algo that is not memory hard? qrk, x11, single algos ...sha 256 ... all the others...

so for asic resistance - - scryptN, vs scrypt jane - high n, vs QRK, vs X11. Let's do that one now.... who wins the anti asic battle and why? technical difficulty to create and cost to create...lets take in all the factors here....??

You could make it run much faster on a farm with a smart paralleling solution.
That would be a thread dispatcher for the algos among entities delivering a timely result according to the total cycle time.
Get a quote from the guys who do the weather forecast software.
Or from Disney, or lets call a name, PIXAR.

Or Google.
Whats up with the Google Coin. If it's for real I'd expect a real overcast for anything Altcoin.
They probably come up with their own algo.

Or maybe they'll just use X11, bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha bwahaha.

They may well use x11 but before they decide to they just need answers to a few of these questions.

1. is it more efficient than qrk?
2. is it more secure than qrk?
3. is it more efficient than scrypt? or is the miner just crippled and can't use the full potential of the card?
4. are there more efficient miners already out for x11 that are more optimised?
5. is it more asic resistant that qrk?
6. is it more asic resistant that scyptN , scrypt jane
7. is it x12, x13, x99 going to be better? will we need to fork all of the coins over and over again?

just yes or no with some kind of reasoning they can follow.... it can't be that hard can it to answer a few questions to the best of your knowledge?

If we had a direct answer to some of those backed up with some real data that could be a great starting point ...

Here beggar, just to dismantle your house of crap I'll give you some details about the facts you don't understand.

1. is it more efficient than qrk? - There is nothing like efficiency to the algo itself. Software could be efficient. The algo is the algo.

2. is it more secure than qrk?

Yes

3. is it more efficient than scrypt? or is the miner just crippled and can't use the full potential of the card?

It does use 100% of the intruction set of available GPUs at a very good level of optimization.

4. are there more efficient miners already out for x11 that are more optimised?

No.

5. is it more asic resistant that qrk?

Yes, it uses 11 algos, Quark only uses 6. It could change to anything else tomorrow.
Fear mongering of "FORKING" is just a matter of deployment of the right software.

6. is it more asic resistant that scyptN , scrypt jane

Of course. The only advantage these methods have is an increased amount of memory necessary to run the algo.
That means higher power consumption and extreme utilization of the hardware, for nothing actually.
It doesn't improve the security of the challenge, it just takes more power to solve.
So it's, in my opinion, a step in the wrong direction, as is the assumption to make a coin more secure by the size of the network.

You`re trying to sell camel shit on a camel market here.

7. is it x12, x13, x99 going to be better? will we need to fork all of the coins over and over again?

No, I think X123 will be the final solution.

Hmmm interesting...let's take this step by step... you don't give clear answers so i will try to clarify what you are saying.

1. No x11 is not more efficient than qrk. Yes we appreciate the efficiency is due to the mining software and the gpu.

2. Is x11 more secure than quark ? you say yes with no explanation.... expand on this. Secure against what attacks? and how is it more secure?? (answer unclear at this time)

3. In your opinion with some justification the miner may not be all that crippled. Perhaps you are correct we don't know. Really 50 % gpu only and that is fully optimised?? Maybe there is a serious bottleneck somewhere and further optimisation is not possible... but i guess nobody knows for sure yet.

4. wow NO... as simple as that? really, explain how you can guarantee this fact? we do need some reason to back up opinion, perhaps you have a good reason but really you didn't express it all that well.

5. Hmm does having more algos essentially make it more asic resistant. Does the fact these are sequential and therefore predictable not make it easier for asics than qrk since that is random? I have no idea on this.. perhaps others can expand on this one. Either way are both x11 and qrk less asic resistant than memory hard algos?

6. No it is not more asic resistant that scrypt N. Since it will be most costly in terms of hardware to implement scrypt N asics. So it seems scryptN is more asic resistant due to costs? or you consider the costs of the extra hardware not as much of an issue as creating an asic to tackle the 11 algos?

7. x123 ....... should we fork to this immediately or build up one at a time ?