Scrypt vs sha256

seems to me that most new coins are with scrypt. is there any advantage over this or is it just asic protection (for the time being) ?

As far as I know, scrypt’s only advantage is its memory-hard problem that it introduces. This means that for the computation done to generate hashes for mining, it also needs a lot of memory.
I think the only reason scrypt is said to be ASIC resistant is because GPUs are already pretty well designed for this application, and if there are going to be scrypt-ASICs (which I believe will happen sooner or later), they will still be not be as much of an improvement on our personal hardware that we use now (for scrypt hashing) that we see in the bitcoin network nowadays.

The reason I wanted to post this is because I think that even though Scrypt’s memory-hard problem proves to be resistant to ASICs today, its real advantage is that the amount of energy consumed worldwide as a result of competition is limited due to the hardware architecture that is required. For someone to be effectively competitive, it will not only mean that he/she needs a lot of processing power, but also means h/she needs a lot of memory. Thus, budget will dictate a balance were less computing power is allowed.

There are a few advantages that I could make up from this.
I personally think that it is a more energy effecient method for proof of work, because of the fact that a smaller part of one’s resources consists of pure prossessing power, meaning less energy (or electricity) needs to be consumed by a miner to mine coins than if sha256d would be used.
Another small advantage is probably that the difference between personal computers and specialized hardware will not be as huge as seen on the bitcoin network. However, there is nothing to prevent people from setting up farms, and farms owns PCs.

Hope that helps?

I’m more attracted by mining Scrypt-based cryptocurrencies than hashcash because I hope that we won’t burn the Earth down just to support an infrastructure. ^^

But your GPU uses more energy than the ASIC’s … so you should love SHA256 more !!!

You don’t want to burn the earth… then you are not a fan of POW mining in general. You should be a massive fan of Peercoin then :slight_smile: as it uses POS minting and this does not require any extra hardware to be running so any old PC will do… the more efficient the computer the more energy efficient your network…

Please read and educate yourself before breeding FUD about scrypt and SHA256 mining. Just because you “feel” you missed the early adoption boat on bitcoin mining does NOT make any scrypt coin better in any way, or energy efficient, or give you a chance in earning more coins per unit of electricity.

Fuzzybear

Yes, of course, ASICs will always be more efficient than GPUs, that what they were designed for!
But my point is that when the time comes that Scrypt ASICs arrive at the market, they’d still use less energy than sha256d ASICs. However, that is assuming that the same amount of resources where used to for building both ASICs.

I thought Peercoin was a proof-of-stake/proof-of-work hybrid? It is indeed interresting, but I haven’t had the time yet to look into it more. Anyway, it was kind of an overstatement, but yes.
I’m also wondering if there is (going to be) a coin that uses Proof of Space? That would be interresting as well.

What? I just concluded that I’m more attracted to one than the other, I have no idea how you could deduce any of that from the few opinions that I’ve stated.
Never have I said I missed ‘the boat’, and I don’t care really. However, I’d say it does give you a chance in earning more coins per unit of electricity, but that is a simplified way of looking at it (there are more factors involved).

I’m not defending Scrypt because of a missed boat, if that is what you think. I’m defending Scrypt because I think it is a technically better alternative, if power consumption is of any concern.

i think scrypt has the advantage it can achieve a larger dispersant of coins than SHA

SHA is at the mercy of asics, and CPU bot nets and server farms.

Scrypt however is everyman and his video cards.

There are not alot of Bot nets or server farms that have the graphics cards required, and of course asics are out.

For US though in PeerCoin it matters less and less as Minting goes up, mining goes down

SHA is at the mercy of asics, and CPU bot nets and server farms.

That seems contradictory. With the proliferation of ASIC and GPU mining for SHA it makes CPU mining even more irrelevant.

[quote=“jubalix, post:6, topic:964”]SHA is at the mercy of asics, and CPU bot nets and server farms.

Scrypt however is everyman and his video cards.

There are not alot of Bot nets or server farms that have the graphics cards required, and of course asics are out.[/quote]

While it isn’t very efficient (for the machine that is mining it), it’s significantly easier to mine Scrypt-based coins with a CPU, so most of the botnets you see today are configured to ghost-mine for Litecoin. A GPU is a must if you are looking to personally mine coins, but if you’ve got access to a couple thousand zombie machines, you’re mostly concerned with quantity over quality of devices, and with (effectively) zero energy costs, there’s no concern with how inefficient your system is, as long as it can stay under the radar and not alert the hosts.

Here’s a recent article about a Dogecoin botnet.

Scrypt coins can, rightly, claim that their hashing scheme is more egalitarian and accessible to hobbyists, today, but I don’t expect that this will last very long. Once the Viper scrypt-ASIC miners (or a competitor) are released, the same “arms race” will affect scrypt coins that currently makes SHA-256 a “big miner game”. It doesn’t matter if the release happens this year, or next, the end result for the network will be the same.

If anything, “hard memory” schemes, like the one that the Ethereum team is proposing in their “Dagger” algorithm will be a step in the right direction. Also, I don’t know if it is is technically possible to come up with a multi-scheme approach (SHA-xxx + scrypt + hard memory = combined hash that is tested), but I’m sure people have thought about it or are working on something similar.

Ultimately though, it doesn’t matter how difficult the algorithm is made or how hardened/difficult/costly it is to produce application specific integrated circuits purpose built to solve the puzzles – if someone has the resources to buy hardware, there’s always going to be “big” miners vs. hobbyists. Just look at some of the insanely large GPU farms that people are building to scrypt mine for an example of this (pictures).