Have the scrypt asics arrived already?
-
I have been watching it…
If you notice there are Pool hoppers that switch pools based upon coin profitability.
If you watch just before (about 1 hr before difficulty changes) you will see the hash rate jump from 4gh to 25gh+! those stupid pool hoppers are taking advantage of the 9% +/- difficulty change and solving blocks in 30 seconds or less… then just before it re-calculates they jump off causing a minor change in difficulty. They are milking the fact that since this is “averaged” out over a certain number of blocks it only slightly raises the difficulty so they can do it again at that time.
Its not a flaw with feathercoin, but rather a vulnerability because they are taking advantage of the system and how it functions.
-
If what your saying is true, this needs to be sorted out.
I think it is in the process though… We are looking at finding an alternative to ACP which should hopefully fix the re-targeting times right? Like a modified version of KGW?
-
Wrapper and Wellenreiter are working on an improved difficulty retarget system that is going to overhaul what we have. It will not act as a replacement for ACP. Wrapper is very keen to find a replacement to ACP but is focused on the immediate problem of the multipools. Version 0.8.6.1 will have a solution to this issue.
On ACP, there are Scrypt ASICs on the horizon and we may well want to move to a new hashing algo so we are still viable for GPU miners and not just those with deep pockets. ACP will be much needed while we transition between hashing algorithms as from one block to the next all the hashing power will be rejected until everyone fires up and mines using the new algo. We would have been incredibly exposed at that point if not for ACP.
-
Wrapper and Wellenreiter are working on an improved difficulty retarget system that is going to overhaul what we have. It will not act as a replacement for ACP. Wrapper is very keen to find a replacement to ACP but is focused on the immediate problem of the multipools. Version 0.8.6.1 will have a solution to this issue.
Ah excellent. Good to hear.
On ACP, there are Scrypt ASICs on the horizon and we may well want to move to a new hashing algo so we are still viable for GPU miners and not just those with deep pockets. ACP will be much needed while we transition between hashing algorithms as from one block to the next all the hashing power will be rejected until everyone fires up and mines using the new algo. We would have been incredibly exposed at that point if not for ACP.
Oh ok. So we are actually considering changing algo’s? I was in support of the idea but was given the impression that it would cause to much grief and wouldn’t be so bad if we stuck with the current algo.
What do you think we could change to? Scrypt with adaptive n factor?
-
It will be interesting where everyone will be pointing their GPUs when the time comes.
-
Wrapper and Wellenreiter are working on an improved difficulty retarget system that is going to overhaul what we have
Yes! Do they accept donations! I’ll gladly lob some spare funds in a pot to pay out on completion of set milestones. :D
-
This topic came up early on in the Feathercoin forum and was a popular suggestion. We launched on the notion that SHA-256 ASICs were about to appear and Scrypt was going to be very important. This captured a lot of peoples attention and after we saw a huge increase in the number of Scrypt coins. Litecoin did very well when the ASICs came out, it become known to be the new home for GPU miners and little by little people shutdown their Bitcoin mining operations and moved over to Litecoin. We could become the new home for GPU miners, right now we compete with Litecoin for hashes and I expect we would fair better in a new tier.
Right now Scrpyt is still in the realm of the GPU miner so the hardware is affordable. If Scrypt becomes dominated by ASICs then only those who can afford £5k+ systems can remain in the mining game. We do not want to be dominated by those with deep pockets in my opinion and remain viable for users of all sizes.
Solutions are not just limited to Scrypt-Jane, there is also mixed hashing solutions out there like DarkCoin and Quarkcoin. We need to look at the effectiveness of this alternatives. Common sense tells me that mixed hashing solution would be more ASIC resistant but I may be wrong.
-
I’m not against Scrypt ASIC, it’s progress, I’m pro progress, but I agree, that the step for entry can’t be set out of newcomers reach. Do ASIC have to become priced out of the reach of the general public? Is it a maximising profits and supply and demand that drives the price up? Is there no way to get involved and keep the price fair and accessible, like Raspberry Pi has and even Google has done with the Nexus series?
-
If Scrypt becomes dominated by ASICs then only those who can afford £5k+ systems can remain in the mining game. We do not want to be dominated by those with deep pockets in my opinion and remain viable for users of all sizes.
I strongly share this belief. The whole point of crypto’s was that it was supposed to be accessible (mining wise) to the average user. Like I mentioned, I was convinced otherwise on the grounds that ftc would be worth more thus little hashing power would still yield the same amount of “profit” (fiat-wise).
With that said, we are missing out on the hundreds of thousands of new miners that are becoming aware of crypto’s. If we want to continue the fundamental aspect of decentralisation, then changing algo’s in a way that GPU miners can continue, but lock out the asics for just a little longer, using a new algo can be a way of achieving this.
I don’t really know about the challenges of doing so, but I really do think by keeping gpu mining active but preventing the asic’s we can keep ourselves as decentralised as ever. Next to preventing pool hoping, this could really shake things up a bit.
Maybe once scrypt asics are affordable the end user, changing back to plain scrypt would be the way to go. I think we should stay ahead of the game and stick to the core fundamentals of crypto-currencies.
Solutions are not just limited to Scrypt-Jane, there is also mixed hashing solutions out there like DarkCoin and Quarkcoin. We need to look at the effectiveness of this alternatives. Common sense tells me that mixed hashing solution would be more ASIC resistant but I may be wrong.
Does darkcoins’ and quarks’ algo allow gpu mining currently?
-
Darkcoin has a GPU miner already, we want to make sure that GPUs are viable on whatever solution we choose. CPU only coins have proven to be dominated quickly by those with access to virtual platforms with hundreds of vCPUs. I’m not sure about Quarkcoin, last I heard they had a miner but it was not that stable, I expect that this has changed.
I expect the pools will hate us for making this move though. I have not spoken to Coinotron but I got the impression he is not always keen on change as he has to get up and do some work :)
-
blake, bmw, groestl, jh, keccak, skein, luffa, cubehash, shavite, simd, echo
Hrm… Interesting. So this is the set of algo’s darkcoin uses? And they have a gpu miner…
In regards to pools having a sad, if we give them plenty of time, surely they would understand.
Anyways. Kewl.
Once the re-targeting thing is sorted, we could look at changing algo’s and this will bring power back into the hands of the individual.
-
Iam totally against switching our algo. Asics arent bad, its an investment pointed direct towards crypto. If you cant afford them, mine other coins or invest into the coin of your choice directly.
I dont see any point to switch. Look what Asics did with Bitcoin the value did go through the roof.
-
Yes it’s value went up, but it centralised hashing… 90% of the little guys bailed btc for ltc, then all the other scrypt’s
Isn’t the point of crypto’s is that everyone gets to process transactions?
In their current form, asics put the power of many into the hands of the few. It’s the reason why fiat is screwing the world over.
-
I think a switch to an already existing algo like scrypt-jane would kill Feathercoin. If we switch, (What I dont hope) a new algo is needed. So better stay @scrypt.
Scrypt Asics arent that much powerful compared to GPU in terms of total hashing power.
-
So maybe we come up with our own brand new algo and precompiled gpu miner.
-
I believe that changing the algo could be very beneficial,
it will avoid first a monopolization of mining where only people with very
expensive hardware could mine
it would avoid altcoin to become only a speculative bubble where
asics miners switch from coins to coins to generate profit
it would be very easy for a pool of ascic miners to control any new coins created
In my opinion if we want feathecoins or any other altcoins to become stronger an adopted as a new way of payment lets make it available to the most not to a minority.
-
Hopefully I didnt stir up a hornets nest!
Next time the difficulty target is about to change (about 1-2 hrs) check out the hashing page and you will notice it spike like crazy and reset the difficulty.
-
Iam totally against switching our algo. Asics arent bad, its an investment pointed direct towards crypto. If you cant afford them, mine other coins or invest into the coin of your choice directly.
I dont see any point to switch. Look what Asics did with Bitcoin the value did go through the roof.
ASICs may have helped Bitcoin in terms of monetary value but look what they did to Terracoin. This is my concern and I do not think that we should talk about price when thinking about what is best.
I have been watching it…
If you notice there are Pool hoppers that switch pools based upon coin profitability.
If you watch just before (about 1 hr before difficulty changes) you will see the hash rate jump from 4gh to 25gh+! those stupid pool hoppers are taking advantage of the 9% +/- difficulty change and solving blocks in 30 seconds or less… then just before it re-calculates they jump off causing a minor change in difficulty. They are milking the fact that since this is “averaged” out over a certain number of blocks it only slightly raises the difficulty so they can do it again at that time.
Its not a flaw with feathercoin, but rather a vulnerability because they are taking advantage of the system and how it functions.
Back to the point, I think you are right. This could be seen as a vulnerability and requires a hard fork to fix. Wrapper has come up with settings which result in far less loss to mining pools and attackers. I expect that we are being attacked with resources from Litecoin, moving hashing algo protects us from large hostile Scrypt pools.
-
The changeover would be a lot of trouble. Everyone will have to get the new miner and start mining with that immediately after the hardfork. There will also be a lot of complaints from miners later on who will have been mining on the old chain.
-
It would not be easy. We would have to give at least a months notice and send the clients constant messages to upgrade. Right now I’m not even sure that we have a viable solution to move to, we do not just need to replace Salsa and SHA-256 we also have to come up with a working GPU miner. There are many other challenges that we need to get past.
Hopefully these miners will eventually thank us for providing a top tier coin that can be mined with GPUs. We could be the only coin on BTC-e that can be mined on a home computer without an expensive ASIC. I understand the concern, this is a brave move but one that needs to be looked at long term, not at the immediate obstacles.
