[XPM] Working on a GPU miner for Primecoin, new thread :)

[vinne81]

So, What are the guesses with how long it'll take before the miner will actually mine? Anyone got a guess to throw up there. My guess is 2 more months.

Never unless someone else picks up the project which mtrlt has seemed to indicate he is abandoning,

Yups, also Primecoin's 15minutes of fame seems to be over. Price will never recover.

Why will the price never recover? Since the threat of a sudden increase in production has been removed, I would expect the price to climb. Or are you suggesting that no-one will want to buy Primecoin?

That is exactly what I am suggesting. Primecoin has a fantastic new mining mechanism, using prime chains. But really that's just a novelty.

Price might fluctuate, but it will never ever go back to 0.01

But that's just, like, my opinion man 

[1714143680]

1714143680

[1714143680]

1714143680

[Asphyxiated]

Yeah, please release something that is able to find blocks and doesn't crash after 5sec-1min.

lol its really not THAT bad, I compiled it on linux and it runs just fine 24 hours a day ( although I only let it run for 25 hours total and then decided it was shit, it did run the -whole- time, not restarts, about 60 million "primes" per sieving session) but i guess windows could be worse, but thats your operating system, not really the miner )

On another note, as I said I would I am implementing x.pushthrough support for ReaperPrime so that it can be hooked up to the existing pools, which is hopefully drive more people to be interested in working on the source code and therefore the miner will be working correctly much quicker than if left as a solo piece of software. At least I hope thats the case, I could do all this for nothing.

[mechs]

Abandon Thread!

Consider this thread abandoned!

It was fun while it lasted.

Maybe we should offer a new bounty, held in escrow, which will be released to a new developer (or team) which can improve this miner so it meets certain pre-set goals in terms of performance and stability. We could even put a reasonable time limit that if the pre-set goals are not met, the bounties are returned to the donators.

I would provide a few coins to such an endeavor - we could use the forum moderator Saltyspoon as the escrow or someone else trusted in the community like John K. or Dooglus to hold the funds. 

[ivanlabrie]

Abandon Thread!

Consider this thread abandoned!

It was fun while it lasted.

Maybe we should offer a new bounty, held in escrow, which will be released to a new developer (or team) which can improve this miner so it meets certain pre-set goals in terms of performance and stability. We could even put a reasonable time limit that if the pre-set goals are not met, the bounties are returned to the donators.

I would provide a few coins to such an endeavor - we could use the forum moderator Saltyspoon as the escrow or someone else trusted in the community like John K. or Dooglus to hold the funds. 

That's a productive post...a change from the rest of them in this thread.
Who's in for this?

[ReCat]

Abandon Thread!

Consider this thread abandoned!

It was fun while it lasted.

Maybe we should offer a new bounty, held in escrow, which will be released to a new developer (or team) which can improve this miner so it meets certain pre-set goals in terms of performance and stability. We could even put a reasonable time limit that if the pre-set goals are not met, the bounties are returned to the donators.

I would provide a few coins to such an endeavor - we could use the forum moderator Saltyspoon as the escrow or someone else trusted in the community like John K. or Dooglus to hold the funds. 

I, too, have thought of this. However... it would be VERY difficult to carry out, I did not consider it plausible to happen, so I did not embrace it.

If you do this, I wish you luck in your endeavors and hope you have success with it.

Anyways, I have created a poll to see the collective emotions of the people of the forum in regards to this miner. https://bitcointalk.org/index.php?topic=296780.0

[lucasjkr]

Maybe we should offer a new bounty, held in escrow, which will be released to a new developer (or team) which can improve this miner so it meets certain pre-set goals in terms of performance and stability. We could even put a reasonable time limit that if the pre-set goals are not met, the bounties are returned to the donators.

I would provide a few coins to such an endeavor - we could use the forum moderator Saltyspoon as the escrow or someone else trusted in the community like John K. or Dooglus to hold the funds. 

Simply, you won't get any dev interest, I don't think.

Clearly, whether making a GPU miner that performs up to "par" (par being what forum users expect) is an endeavor. Ask yourself why someone would go through the effort to attempt to claim the bounty, then either find out that someone else beat them to it OR that they did the work to the absolute best of their ability and the performance still isn't acceptable to the people posting hte bounty? Why that, when they could at least reap a much bigger reward by doing it as a completely solo endeavor, and if they proved to be successful, never mention a word of it and simply have a huge advantage over the rest of the network.

Seems lots of people want the fruits of someone elses labor but are unwlling to pay for that labor. That's not going to get you what you want. Sure, there may be disappointments along the way, but such is life.

Mtlrt (forgive the butchering on the spelling) provided exactly what was promised; prior to his efforts, it wasn't even clear if a GPU miner could exist; applaud him for that. But this idea about trying to get a reward in exchange for taking absolutely no risk, well, that doesn't strike me as very equitable for teh people who actually have the skills you're seeking.

[2Kool4Skewl]

Mtlrt (forgive the butchering on the spelling) provided exactly what was promised; prior to his efforts, it wasn't even clear if a GPU miner could exist; applaud him for that. But this idea about trying to get a reward in exchange for taking absolutely no risk, well, that doesn't strike me as very equitable for teh people who actually have the skills you're seeking.

Uhhhh....  you "applaud him"  lol

He knew his software didn't work but he kept collecting money up until the day he "released" it.  You applaud him for this?

Not to mention, he was supposed to provide a working "beta"... How many blocks did his GPU miner find?  I don't think it's even found one.  lol

Anyone that still supports him is either the biggest stooge or a complete idiot.

[CoinHoarder]

The Primecoin community should be happy this was a failure. If I owned any Primecoin, I know I would be. As long as it stays CPU-only, it will stand out from the sea of newly released GPU mined scrypt coins, for purely the simple fact that it can only be mined on a CPU. Forget the fact that it is truly innovative in the terms of the prime number proof of work algorithm.

This allows people to mine Primecoin with their CPU while mining other cryptocoins with their GPUs. If Primecoin GPU mining became a widespread occurrence, this would force people to choose whether to continue mining Primecoin (but on their GPU instead), or to mine another coin with their GPUs.

Keeping Primecoin a CPU only coin increases its user base by allowing more people to participate in mining, whilst mining other cryptocoins with their GPUs. I'm sure Primecoin picks up many users from many different ALT coin communities for purely the reason it is CPU only. Otherwise, they would have to decide whether they want to mine Primecoin or not, and it wouldn't be a "no brainer" like it is now. It is obviously the most innovative CPU-only cryptocoin, if it goes to GPU then it is not so much obviously the most innovative (or lucrative) GPU mine-able coin that it is while it remains CPU only.

Don't get me wrong, if I would of donated to the cause I would be slightly pissed at the moment. However, I think this is a blessing in disguise for Primecoin. The longer it remains CPU only, the larger user base it will acquire. I do expect Primecoin to be eventually GPU mined, but again the longer this takes to happen, the better off Primecoin will be.

[YipYip]

The Primecoin community should be happy this was a failure. If I owned any Primecoin, I know I would be. As long as it stays CPU-only, it will stand out from the sea of newly released GPU mined scrypt coins, for purely the simple fact that it can only be mined on a CPU. Forget the fact that it is truly innovative in the terms of the prime number proof of work algorithm.

This allows people to mine Primecoin with their CPU while mining other cryptocoins with their GPUs. If Primecoin GPU mining became a widespread occurrence, this would force people to choose whether to continue mining Primecoin (but on their GPU instead), or to mine another coin with their GPUs.

Keeping Primecoin a CPU only coin increases its user base by allowing more people to participate in mining, whilst mining other cryptocoins with their GPUs. I'm sure Primecoin picks up many users from many different ALT coin communities for purely the reason it is CPU only. Otherwise, they would have to decide whether they want to mine Primecoin or not, and it wouldn't be a "no brainer" like it is now. It is obviously the most innovative CPU-only cryptocoin, if it goes to GPU then it is not so much obviously the most innovative (or lucrative) GPU mine-able coin that it is while it remains CPU only.

Don't get me wrong, if I would of donated to the cause I would be slightly pissed at the moment. However, I think this is a blessing in disguise for Primecoin. The longer it remains CPU only, the larger user base it will acquire. I do expect Primecoin to be eventually GPU mined, but again the longer this takes to happen, the better off Primecoin will be.

Theoretically I agree with you...

The main problems i see are

1) Raped & pillaged by botnets
2) XPM price tanked like a biatch after the GPU failure
3) Would have opened it up to a lot of serious miners

[balanghai]

so now this is long gone? Maybe the developer now has ordered great rig for his bitcoin mining. 

[CoinHoarder]

Theoretically I agree with you...

The main problems i see are

1) Raped & pillaged by botnets
2) XPM price tanked like a biatch after the GPU failure
3) Would have opened it up to a lot of serious miners

I understand your points.

1.) My only rebuttal is that Litecoin was CPU only at its inception and look at it now. I agree as long as it is CPU only, botnets will have a huge advantage due to economies of scale. However, you could say the same about GPU mined coins... every computer also has a GPU, there is no stopping a botnet from utilizing them. Both Bitcoin and Litecoin were and/or are currently mined with botnets even on GPUs. If Bitcoin and Litecoin can survive an era of botnet domination, then so can Primecoin IMO.

2.) I would not worry about short term price fluctuations. As you know, cryptocoins are highly volatile and cryptocoin traders are largely inexperienced. Most of them suffer from herd mentality, they see the price falling so they say to themselves "oh no, I better get rid of it!" I see good things in Primecoin's future for the simple fact it is currently cpu only, not to mention the innovation and its contribution to the field of mathematics. In the long run I think my opinion on this will be proven correct, but we shall see. Guess which other coin started out as CPU for a while, then came to GPUs and was a big success? Hint: It has risen over 2500% in the past year.

3.) I agree it would have opened Primecoin up to more serious miners... I have no rebuttal here. Well.... actually I guess I do. mrtlt has proven it is possible to mine Primecoin on GPUs. It is now not a matter of if Primecoin will be GPU mined, but when it will be GPU mined. For now it stays off limits to large GPU farms, but the day will come (likely sooner rather than later) when that is no longer true. Again, if Litecoin can make it through the same scenario, there is no reason Primecoin can't as well.

Wow.. I have seriously just talked myself into buying some Primecoins in my last two posts. I honestly do not own any at the moment, but I am actually kind of excited about an ALT coin for the first time in a while. That is saying a lot, because I have been less than impressed with just about every ALT coin release this year.  

[Trillium]

The main problems i see are

1) Raped & pillaged by botnets
2) XPM price tanked like a biatch after the GPU failure
3) Would have opened it up to a lot of serious miners

Ugh, typed up ~1 page response. Database error when posting. I need to remember to CTRL+C large posts before submitting them...

quick version:

1) Any real evidence of this aside from a few AV companies blacklisting the code? Same can be said for bitcoin and also all the other scrypt altcoins that are worth a damn.
2) price tanked well before the alpha/beta release to donors. had nothing to do with news of the failure because the failure came much later.
3) Not really, it was always going to be a GPU accelerated miner not a GPU dedicated miner. Most litecoin rigs have poor CPU and would have struggled. Made worse by their outrageous CPU:GPU performance ratio mis-match. I think even if it spread amongst existing scrypt miners who became interested in primecoin, they'd still be washed away by the VPS miners who rent VPS CPU by the hundreds or thousands. I think too many people thought that the GPU miner release would be like when bitcoin got its first GPU miners. In reality the mathematics of primes and SHA-256 are nothing like each other and the hardware required is not the same. This should be evident from the GPU miner project failure.

[ReCat]

If CPU mining is the only way forever, then you will face problems with huge botnets mining, and then maybe even 51%ing the network. With a GPU miner, the difficulty will rise sufficiently so that only specialized miners can take control of the network, specialized individual miners, making it more difficult to 51% the network because when it's individuals, you can't come to an agreement to 51% like that.

[Trillium]

If CPU mining is the only way forever, then you will face problems with huge botnets mining, and then maybe even 51%ing the network. With a GPU miner, the difficulty will rise sufficiently so that only specialized miners can take control of the network, specialized individual miners, making it more difficult to 51% the network because when it's individuals, you can't come to an agreement to 51% like that.

Well primecoin has been out for a few months now and its difficulty is currently very stable. It has been precisely 9.86 +/- 0.01 for a week or more (I think, from my manual checks. All the good graphing/data sites seem to be down or no longer updating).

If botnets were growing day in and out we'd see diff rising but we aren't. It's unlikely that botnets would be growing and other users would be dissapearing at exactly the same rate so I don't think that is a likely possibility either.

[mhps]

All the good graphing/data sites seem to be down or no longer updating).

These charts http://xpm.muuttuja.org/charts/  are up to date and Mikaelh's analysis at http://www.ppcointalk.org/index.php?topic=491.0 is interesting.

Move the left bar below the time-axis to magnify recent changes.

I think the combination of dropping XPM price and dimming prospect of the GPU miner makes a lot of mining power to hold fire, or simply pack up and leave. The most recent increase starting on Sep. 14 was from the release of hp11.

[Trillium]

All the good graphing/data sites seem to be down or no longer updating).

These charts http://xpm.muuttuja.org/charts/  are up to date and Mikaelh's analysis at http://www.ppcointalk.org/index.php?topic=491.0 is interesting.

Move the left bar below the time-axis to magnify recent changes.

I think the combination of dropping XPM price and dimming prospect of the GPU miner makes a lot of mining power to hold fire, or simply pack up and leave. The most recent increase starting on Sep. 14 was from the release of hp11.

Thanks for the link it's exactly what I was after.

[ReCat]

If CPU mining is the only way forever, then you will face problems with huge botnets mining, and then maybe even 51%ing the network. With a GPU miner, the difficulty will rise sufficiently so that only specialized miners can take control of the network, specialized individual miners, making it more difficult to 51% the network because when it's individuals, you can't come to an agreement to 51% like that.

Well primecoin has been out for a few months now and its difficulty is currently very stable. It has been precisely 9.86 +/- 0.01 for a week or more (I think, from my manual checks. All the good graphing/data sites seem to be down or no longer updating).

If botnets were growing day in and out we'd see diff rising but we aren't. It's unlikely that botnets would be growing and other users would be dissapearing at exactly the same rate so I don't think that is a likely possibility either.

of course. it just remains a possibility.

[Stinky_Pete]

Well primecoin has been out for a few months now and its difficulty is currently very stable. It has been precisely 9.86 +/- 0.01 for a week or more (I think, from my manual checks. All the good graphing/data sites seem to be down or no longer updating).

If botnets were growing day in and out we'd see diff rising but we aren't. It's unlikely that botnets would be growing and other users would be dissapearing at exactly the same rate so I don't think that is a likely possibility either.

+1. Just quietly priming away in a quiet and stable environment.

[cycloid]

Buying up all the cheap XPM ... this is gotta be the best deal you will ever see on this coin, yey I feel very lucky to get it this cheap ! 

[wetroof]

botnets are a problem but it depends how many. If botnets only make up 5% of the network, then they are having small effect on the price - and the price is mostly being set by honest miners.

the same can be said for server administrators taking advantage of their networks, mining primecoin at zero expense. There is only a finite capacity of botnet miners and zero-expense miners. I think they can be overwhelmed in time by serious miners who set up professional cpu rigs - who then becomes the major determiners of the xpm price.

At this point I think at-home-miners are struggling to overtake the zero-expense miners. and probably will for some time, unless more investments are made into xpm mining.

and GPU mining is still a possible factor... which might make investment into cpu-rigs cautious.