Topic: [XPM] Working on a GPU miner for Primecoin, new thread :) - page 10. (Read 166591 times)

Never mind, an old version of libblkmaker was being kept by the package manager, it works now

I was wondering why I can't use this for the beeeeer.org pool because it uses getwork just like reaper does but if i point it at the ip address of www.beeeeer.org it give me a curl error 52 unable to get work......

Anyone else tried this at all?

The paying vps are already under the water at current price. If working GPU miners do not appear, the non-paying or little-paying vps and botnet will donminate. The non- or little-paying vps are mostly from university and research/government computer centers where mining is basically fleecing. A working GPU will not only let current  individual XPM miners increase their mining power, but also draw an army of scrypt miners to XPM mining and probably overwhelm the current XPM miners.  That will drive difficulty up, reduce reward, and put a floor on the price. However I am not saying the price will go up immediately after a working GPU miner release. Short-term price action is difficult to predict.

I am having trouble getting reaper to accept a different primecoin address. If i leave the primecoin_address setting to the default address everything runs just fine but I assume that I wont be getting paid if i were to find any blocks.

The libblkmaker i have installed is the one that was patched for primecoin so I dont know why it will not accept my primecoin address but will the other one....

thanks

Saying that you need to assume that GPU mining Primecoin is not only significantly faster than CPU mining, but more energy efficient as well (otherwise you can consider it an increasing number of CPU miners). First steps in implementing GPU mining software allow to hazard a guess that the difference between CPU and GPU mining primecoins might not be as significant as it was for the cryptocoins that have hash based proof-of-work approaches.
But even if a GPU miner is faster, the rewards per block will react to that increase of network calculating power. So you might add more power to the network, but in difference to other cryptocoins, where this allows you to get bigger parts of the constant rewards, Primecoin "punishes" you for hugely increasing the calculating power with decreasing rewards.

That leads me to the assumption:
it is hard to predict the influence of GPU miners for Primecoin. In my humble opinion they will have a lesser impact than they had for other coins - in terms of increasing the calculating power. So I assume, the price will not be very dependent on GPU miners. And even if the GPU miners brought much more calculating power to the network, that reduced the amount of generated coins over time which could have a stabilizing effect on the price.
I don't see that huge drop coming. But my crystal ball is under maintenance so I can't tell for sure...

I just want to add a rough estimation.
Say the GPU miner increases tenfold the calculating power compared to a CPU miner (I'm not saying anything about GPU/CPU types, models, etc. ).
Say the GPU miner needs double the electricity as the CPU miner for that tenfold increase.
If you are the one and only with such a GPU miner you have a fivefold energy efficiency and a tenfold reward per time.
But as soon as others draw level, something happens:
compared to the current difficulty (now) of approximately 9.86 a complete replacement of CPU miners by GPU miners would push the difficulty to roughly 10.86 (assuming that the tenfold increase does this, but afaik this is the logarithmic way how the difficulty works).
The result for the rewards per block would be a decrease from 999/(9.86)^2=10.27 to 999/(10.86)^2=8.47
From that moment on the energy you spend is still double, but the rewards have dropped to 82% of what they have been before (I know it is much more dynamically, but I try to create a scenario ).
Either the price per Primecoin rises or the miners have to consider bailing out - if they don't want to invest more money for buying electricity than they had to for buying the same amount of Primecoins at an exchange (and buying at an exchange rather than mining would be a rational consequence!)
So I (once more without crystal ball and without engagement) expect the Primecoin price not to drop that much just by the rise of GPU miners (based on the aforementioned assumptions).

Well, I predicted XPM drop more than a month ago when this miner was announced, see the last sentence of this post:
https://bitcointalksearch.org/topic/m.2876322
Then sold all few hundreds XPM I mined in the XPM beginnings for 0.0062 in fear of GPU miner. And I was right, I can buy them back today for almost half that price, but I will not because I still think that working version of GPU miner will come out and XPM may go to 0.001 and than I may buy some back. Logic behind it is still the same: If you do VPS you have to sell now and sell expensive, if you GPU mine you can dump for a fraction of the price, what do you care how high VPS costs are? As long as VPS get out of the game there is more XPM to be mined for you. So XPM price has to go down significantly in the proportion of GPU miners ability to kick VPS servers out.

Working GPU miner, when released to the public, will drive XPM price up. At least, that is my biased opinion

They could hold the coins until the prices raise if they have the financial wherewithal to do so.  Still, for the majority of VPS miners, you are probably right.  A good GPU miner by both increasing difficulty, network strength & decentralization would probably lead to higher prices.

At the end of every month at the latest to pay hosting providers or you are shutdown. Not exactly like when you have a machine with a spare GPU which can crank coins with minimal costs, than you can choose if you will sell them or not. With VPS you have no choice but sell.

So VPS mined coins are to be sold immediately, aren't they?

Exactly the opposite. VPS generated coins are high-cost coins, hosting + electricity has to be paid, no mercy. Single guy with his 4x-Radeon mining rig has to pay only electricity, all other costs can be postponed, so his coins are low-cost. Price will sink like a rock at the very moment somebody successfully debugs this miner.

It does. A working GPU miner will drive the difficulty up and vps and botnets away. When there are not a lot of low-/no-cost coins being generated the dumping will stop and the price will stablize if not go up.

Because BTC is back up to $140+ USD/BTC. Keep your panties on.

Price is dropping like a rock.  I don't think it matters anymore.

Please, put your pitchforks away...

The miner looks totally legit.  If mrtlt came up with a bogus miner, then he spent a lot of time making it look like it works.

I saw what looked like a legit block solution that didn't fail the primality or difficulty tests, rather some early packet size filter that caused the submission to get rejected.  Sounds like a bug somewhere, bound to get addressed.

If the rumour mill is correct and the mrlt realsolid coinhunter senario is true ...

MicroCash 500BTC & nothing 
GPU miner ~100btc & nothing that works

These are the recent "DONATIONS" operations he has pulled in the last 3 months 

Unfortunately, mr.Mtrlt has a tendency to look really shady and scammy for no reason at all.

 

The source was with the leak.  Has no one uploaded it to github yet?

Its been 2 weeks since the GPU release fiasco & so far crickets

Well to be honest mrlt aka realsolid aka CoinHunter has taken the money and said a nice big FUCKU all to the community after $10k

Has he even released the source so others can finish his abortion in a bucket...?

+1