MMC GPU mining is super profitable!

[tbear]

If it only was possible on geforce cards i would be soooo happy

Only ati

https://bitsharestalk.org/index.php?topic=1784.msg20058#msg20058

you can

[ManeBjorn]

That is a good breakdown.
Have you also tried using YAM miner for MMC as it gives you even better performance?

I did my own CPU mining tests with a bunch of different CPU coins and CPU coins that can also be mined with a GPU here are my results,
http://www.reviewoutlaw.com/most-profitable-cpu-coin-list-alt-coin-profits-24-hours-mining/

MemoryCoin is still winning as of 2/22/14 but that may have changed by the time you read this so give it another look

[iwantcoins]

I did my own CPU mining tests with a bunch of different CPU coins and CPU coins that can also be mined with a GPU here are my results,
http://www.reviewoutlaw.com/most-profitable-cpu-coin-list-alt-coin-profits-24-hours-mining/

MemoryCoin is still winning as of 2/22/14 but that may have changed by the time you read this so give it another look

[reorder]

i'm new to this and don't have a lot of insight to offer, but i'm sticking with mining PTS because i'm uncomfortable with the GPU situation with MMC. but CPU mining is also the entire reason i got into PTS, and then my interest in PTS led naturally to MMC

It is not like there are no GPU miners for PTS out there.

Yeah, I wonder why noone created a pool with that one.

I personally was too late to the game with my miner to the point where selling it was a way quicker reward on the hard work hours spent. And received a pretty lucrative offer on that, I must admit.

[5chdn]

i'm new to this and don't have a lot of insight to offer, but i'm sticking with mining PTS because i'm uncomfortable with the GPU situation with MMC. but CPU mining is also the entire reason i got into PTS, and then my interest in PTS led naturally to MMC

It is not like there are no GPU miners for PTS out there.

Yeah, I wonder why noone created a pool with that one.

[FreeTrade]

Well, why not, one can use 'sea of hashers' technique they use in some Bitcoin ASICs where there are many identical circuits each calculating one round of SHA. If you achieve the same gigahash throughput they have in Bitcoin, storing/loading a hash in RAM becomes slower than just calculating it 50 times on average.

Well not much slower I hope. Let's not forget the 50 rounds of AES too. I think a ASIC designer really has his work cut out for him here.

[reorder]

In fact, there is an obvious shortcut in the algorithm that could make FPGA/ASIC implementation feasible: you do not necessarily need to store all those SHA hashes in RAM and can rather calculate them as you go. Though, the number of required gates to do it reasonably fast may exceed what modern FPGAs provide. Not sure on this one, perhaps someone proficient in hardware engineering could do a rough estimate?

Hmmm - can't believe it could ever be faster to generate them on the fly every time they are needed rather than generate them once and read them in from memory-  even with the most specialized of equipment. But interested to hear any specialist views on that.

Well, why not, one can use 'sea of hashers' technique they use in some Bitcoin ASICs where there are many identical circuits each calculating one round of SHA. If you achieve the same gigahash throughput they have in Bitcoin, storing/loading a hash in RAM becomes slower than just calculating it 50 times on average.

[FreeTrade]

In fact, there is an obvious shortcut in the algorithm that could make FPGA/ASIC implementation feasible: you do not necessarily need to store all those SHA hashes in RAM and can rather calculate them as you go. Though, the number of required gates to do it reasonably fast may exceed what modern FPGAs provide. Not sure on this one, perhaps someone proficient in hardware engineering could do a rough estimate?

Hmmm - can't believe it could ever be faster to generate them on the fly every time they are needed rather than generate them once and read them in from memory-  even with the most specialized of equipment. But interested to hear any specialist views on that.

[reorder]

It only took 3 weeks for a GPU miner to be released for MMC that has a hashing rate an order of magnitude greater than what is possible with CPU mining. Compare to Primecoin (and Datacoin et al) where after 1/2 year and much greater economic incentives, attempts at producing a GPU miner haven't amounted to much of anything (someone released one, but it was not faster than CPU mining; nobody uses it).

It's counter-intuitive, but a GPU miner is very good news for MemoryCoin. It removes the unknowns. We have a pretty good idea now of what GPUs are capable of compared to CPUs and it's not a deal breaker. With PrimeCoin and other coins claiming to be GPU resistant, usually through complexity, you don't know if
a) the dev has been GPU mining all along
b) private GPU software have been competing alongside
c) whether a GPU miner will be developed at a later date when there is sufficient economic incentive.

We have an algorithm that's simple, that relies on limitations of the hardware, that's been tested and has good potential for long-term GPU/ASIC resistance.

I think what'll you see happen over the next few weeks is that GPUs will drive the difficulty up so that they find profits reduced and so will switch back to other coins.

In fact, there is an obvious shortcut in the algorithm that could make FPGA/ASIC implementation feasible: you do not necessarily need to store all those SHA hashes in RAM and can rather calculate them as you go. Though, the number of required gates to do it reasonably fast may exceed what modern FPGAs provide. Not sure on this one, perhaps someone proficient in hardware engineering could do a rough estimate?

[Delinquency]

How far will profitability be pushed down? CPU profitability is currently 33% per month and GPU profitability is 150% MMC mining. So this will push down GPU profitability to numbers below litecoin profitability -- I believe it is 25% (1/4 of capital cost)
with mmc allowing miners to capitalize using gpu's, 25% gpu profitability would drive cpu profitability to around electricity costs? And that's like 5.8% of equipment for cpu miners.

$400 makes $600 of mmc per month gpu mining.
$300 makes $100 of mmc per month cpu mining.

$400 makes $100 per month gpu mining for litecoins.

[FreeTrade]

It only took 3 weeks for a GPU miner to be released for MMC that has a hashing rate an order of magnitude greater than what is possible with CPU mining. Compare to Primecoin (and Datacoin et al) where after 1/2 year and much greater economic incentives, attempts at producing a GPU miner haven't amounted to much of anything (someone released one, but it was not faster than CPU mining; nobody uses it).

It's counter-intuitive, but a GPU miner is very good news for MemoryCoin. It removes the unknowns. We have a pretty good idea now of what GPUs are capable of compared to CPUs and it's not a deal breaker. With PrimeCoin and other coins claiming to be GPU resistant, usually through complexity, you don't know if
a) the dev has been GPU mining all along
b) private GPU software have been competing alongside
c) whether a GPU miner will be developed at a later date when there is sufficient economic incentive.

We have an algorithm that's simple, that relies on limitations of the hardware, that's been tested and has good potential for long-term GPU/ASIC resistance.

I think what'll you see happen over the next few weeks is that GPUs will drive the difficulty up so that they find profits reduced and so will switch back to other coins.

[reorder]

i'm new to this and don't have a lot of insight to offer, but i'm sticking with mining PTS because i'm uncomfortable with the GPU situation with MMC. but CPU mining is also the entire reason i got into PTS, and then my interest in PTS led naturally to MMC

It is not like there are no GPU miners for PTS out there.

[progmac]

i'm new to this and don't have a lot of insight to offer, but i'm sticking with mining PTS because i'm uncomfortable with the GPU situation with MMC. but CPU mining is also the entire reason i got into PTS, and then my interest in PTS led naturally to MMC

[whynot]

There's no way to lock out gpus completely, the only possibility is making it as hard as possible for them.
PTS and MMC are doing pretty good there, with a maximum of 10x more speed. Have a look at what the difference is for other coins and you will see that's a huge improvement.

It only took 3 weeks for a GPU miner to be released for MMC that has a hashing rate an order of magnitude greater than what is possible with CPU mining. Compare to Primecoin (and Datacoin et al) where after 1/2 year and much greater economic incentives, attempts at producing a GPU miner haven't amounted to much of anything (someone released one, but it was not faster than CPU mining; nobody uses it).

I'm sorry for being negative in this post, but as news about MemoryCoin 2.0 spreads through the larger crypto-currency community there are going to be a lot of people who will be more critical than those in this forum. We need to be prepared for that -- the best way to prepare is to honestly assess the strengths and weakness of MMC.

[seraphim]

There's no way to lock out gpus completely, the only possibility is making it as hard as possible for them.
PTS and MMC are doing pretty good there, with a maximum of 10x more speed. Have a look at what the difference is for other coins and you will see that's a huge improvement.

Developing ASICs for PTS and MMC won't be profitable for a long while, so people can still earn some coins with their CPUs - and that's what it really is about.