Topic: [ANN][YAC] YACoin ongoing development - page 63. (Read 380128 times)

I am going to use one word to describe coin supply. Would you like infinite or unlimited?

not true, there's no coin supply limit. the 2 bil limitation applies only to a single input.
that said, the equilibrium part still holds. it all depends on how the tx volume, pow supply and pos supply balance out.

Regarding the total supply - this is from sairon :

I would have to go look at some other posts on the total supply question.

YACoin uses Scrypt-Chacha (aka scrypt-jane) with increasing NFactor as the mixing algorithm and SHA-3 (Keccak 512) as the hashing algorithm.

to contrast with Litecoin, Litecoin uses Scrypt (salsa 20/8) as the mixing algorithm and SHA-2 (SHA256) as the hashing algorithm.


edit: duh, mix, hash backwards...

See this for PoW algorithm:
http://forum.yacoin.org/index.php?topic=322.msg1495#msg1495


This is current data:

yacoind getinfo
{
    ...
    "moneysupply" : 24557303.97361400,
    ...
}



With current difficulty some 80K new coins get generated every day:

yacoind getmininginfo
{
    "blocks" : 606896,
    ...
    "difficulty" : 0.00710212,
    ...
    "Nfactor" : 15,
    "N" : 65536,
    "powreward" : 57.02000000
}

YACoin is on the list, but I need info.

What is the total supply? That is blank in first post.

What exactly is the PoW algo? One part says Scrypt-N SHA-3, another says just Scrypt-N and in the next sentence talks about SHA-3 Keccak. Finally the miner uses Scrypt-chacha. I know the SHA-3 winner is Keccak but how does it fit into YACoin?

280x 3GB 1.1KH/s (LG 8 R 2560)

R7 240 much better

I was thinking about this more, I will put YACoin back on with the GPU Advantage shown as "researching". I do that with most coins when they get a new GPU miner.

I can do the CPU testing tomorrow.

That's going to put YACoin on the CPU friendly coin list for certain   Please make sure you use the CPU miner optimized for YACoin.  I've got published source and binaries on github at http://github.com/Thirtybird/CPUMiner

I don't recall seeing anyone publish 280x numbers, but as a 3GB card I would guess it's going to be in the neighborhood of 1.5 KH/sec.  If anyone has one and is willing, I'll work with them on settings and we'll see what it can do.

I just calculated $1.97 per day...

It can net 270h/s mining monero, so: http://www.whatmine.com/

While I have been further researching possibilities of improvement I have come up to rather lengthy discussion, after which a proposal has been put just a few days ago:

http://eprint.iacr.org/2014/452.pdf

Could yacoin be upgraded with this in order to help solve checkpointing problem (perhaps even chaintrust?).

It's a lot of material, if you are curious just check "3 Protocol" section in the document and first post of originating thread:
https://bitcointalksearch.org/topic/proof-of-activity-proposal-102355

And I'm getting 410 hash/s on an I7-3770S which equates to 22 cents per day before electrical costs --> www.whattomine.com

I'm wondering how that matches up with other CPU coins on the list regardless of the 'GPU advantage'...

I know it's stupid but I'm curious how much hashrate would you get on stock 280x?

I was wondering if we could use percentage of big PoS miners profit to reward PoW miners that secured PoS blocks in the past. This solution is not attractive yet, because there is not enough money supply.

Here is more or less current information and let's simplify it:

Money Supply:   24M
PoW reward:   57 coins

If there was no PoS, we would get something like this:

PoW currency per day:   82K    (57*60*24)
PoW currency per month:   2.4M   (57*60*24*30)
PoW currency per year:   29M   (57*60*24*30*12)


Maximum PoS reward per year with 24M coins in existance:

24M * 0.05 = 1.2M


Now 1.2M is only half of what would be PoW mined in a month now. So not before money supply hits 600M there will be enough money created with PoS to take over generation rate with PoW.


This is all theory and I know numbers vary each minute (diff, number of PoS blocks...). I am not even sure if 1:10 PoS:PoW ratio means there will always be multiple times more PoW than PoS blocks.

But when we get to that currency ammount (or even before), big PoS miners could start rewarding PoW miners. I thought of this now that I could give a percentage of PoS rewards toward securing blockchain (and that would eventually incentivise PoW miners to secure PoS blocks). If I remember correctly, St.Bit was considering doubling PoW rewards so miners would not ignore PoS blocks for profit sake.
At first I was afraid of this in fear of currency expansion. But that really wouldn't matter when money supply gets big enough for PoS to take over generation.

Big PoS miners could also leverage their reward toward PoW miners that secured "smaller" Pos blocks in between larger PoS-PoW blocks - that would slow down inflation.
In a way this would transfer wealth from the rich but that doesn't worry me since small stakeholders will lose anyway when they don't get their PoS excepted in the chain. On the other hand, rich would get richer.

Could this redistribution be enforced by giving "big" PoS blocks small trust value when they generate a lot of money and they did not include transactions towards older PoW miners who secured previous PoS blocks?
Another way would be to direct a portion of generated interest to those miners.

I was asking this feature long ago! So I'm for it.

I've got some insight for you:

Radeon R7 240 4GB - 1.1 KH/sec
Radeon R7 250 2GB - 0.8 KH/sec

Core i7 3610 laptop - 0.31 KH/sec

No GPU with 2GB of memory or less has shown to hit over ~1.2 KH/sec (750Ti).  I don't have solid numbers on the higher end AMD cards with 4GB, but none of them can utilize all of the available shaders without lookup-gap going to the moon which really kills performance.  If they hit much higher than 2.5 KH/sec I would be surprised.

so, the GPU advantage is likely less than 10x depending on what hardware you choose for GPU and CPU.

Not such yet, at next N change at 5 sep speed will possibly balance chances GPU/CPU.

I believe the spread is quite a bit less than the 10x cutoff for your CPU Coin List.  There aren't a lot of details on your website about exactly which CPU vs. which GPU would be use to make that calculation.  Given that all of my GPUs have insufficient RAM to mine YAC efficiently at the current value of N, I'm unable to provide GPU hash rates, but I'm sure a few people with 3GB and 4GB cards will weigh in with their current rates.

For comparison, note that the CPU hash rate table I posted on the first page of this thread was fairly old hardware, an IBM blade server with 2x Xeon E5450's (each is basically a 3GHz Core2 Quad), yielding about 187 hash/sec at the current value of N.  Majority of desktop PC's have processors faster than that blade server.  A 2x E5450 server has a Passmark score of about 7778:

Dual E5450:
http://cpubenchmark.net/cpu.php?cpu=Intel+Xeon+E5450+%40+3.00GHz&id=1236&cpuCount=2

An i7-2600k was for a long time the happy medium for people to balance price vs performance on desktop PC's, with a Passmark score of 8597, but is getting a bit "long in the teeth" these days:
http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-2600K+%40+3.40GHz

Something like an i7-4770k would probably be more typical today, at a Passmark score of 10299:
http://www.cpubenchmark.net/cpu.php?cpu=Intel+Core+i7-4770K+%40+3.50GHz

So, looking at my old CPU hash rate table, for today's PC's, it would probably be in the ballpark to approximate CPU mining as between 1.1x to 1.32x faster than the table I posted for a typical new PC.  Minor variations will occur due to memory speed and cache configuration in the specific CPU type, of course.

That old CPU hash rate table I posted was at:
https://bitcointalksearch.org/topic/m.2162620