With ASICs on the horizon should DigiByte switch algorithms?

[DigiByte]

What sort of hash rates to you see with X11 & Scrypt-N is it about the same as Scrypt? Or is their a reduction in hash output?

Never bothered with Scrypy-N, But with x11 I get the following:

1.8 mh/s on an Asus 7970
1.24 mh/s on a R9-270

So you are actually getting a hash increase over Scrypt/CGminer?

[sammy007]

What sort of hash rates to you see with X11 & Scrypt-N is it about the same as Scrypt? Or is their a reduction in hash output?

Never bothered with Scrypy-N, But with x11 I get the following:

1.8 mh/s on an Asus 7970
1.24 mh/s on a R9-270

So you are actually getting a hash increase over Scrypt/CGminer?

Different algos, different hashrates... What we are talking about? It does not matter.

[MP5KU]

What sort of hash rates to you see with X11 & Scrypt-N is it about the same as Scrypt? Or is their a reduction in hash output?

Never bothered with Scrypy-N, But with x11 I get the following:

1.8 mh/s on an Asus 7970
1.24 mh/s on a R9-270

So you are actually getting a hash increase over Scrypt/CGminer?

Yep. Different Algorithms so different hashrates

         Scrypt Equiv    x11
7970    720kh/s         1.8 mh/s
270     480kh/s          1.24 mh/s

[DigiByte]

What sort of hash rates to you see with X11 & Scrypt-N is it about the same as Scrypt? Or is their a reduction in hash output?

Never bothered with Scrypy-N, But with x11 I get the following:

1.8 mh/s on an Asus 7970
1.24 mh/s on a R9-270

So you are actually getting a hash increase over Scrypt/CGminer?

Yep. Different Algorithms so different hashrates

         Scrypt Equiv    x11
7970    720kh/s         1.8 mh/s
270     480kh/s          1.24 mh/s

Interesting, thanks for sharing.

Was talking to the Doge devs and they brought up some good points. The stratum on several pools will have to be changed and at the exact block of change we could see some wacky things occur with the net hash rate.

Most mining pools do not support Scrypt N at the moment. Would we be doing more harm than good?

[MP5KU]

What sort of hash rates to you see with X11 & Scrypt-N is it about the same as Scrypt? Or is their a reduction in hash output?

Never bothered with Scrypy-N, But with x11 I get the following:

1.8 mh/s on an Asus 7970
1.24 mh/s on a R9-270

So you are actually getting a hash increase over Scrypt/CGminer?

Yep. Different Algorithms so different hashrates

         Scrypt Equiv    x11
7970    720kh/s         1.8 mh/s
270     480kh/s          1.24 mh/s

Interesting, thanks for sharing.

Was talking to the Doge devs and they brought up some good points. The stratum on several pools will have to be changed and at the exact block of change we could see some wacky things occur with the net hash rate.

Most mining pools do not support Scrypt N at the moment. Would we be doing more harm than good?

Well, depends how you see it.
Moving away from Scrypt will annoy the people with gridseeds and the KNC titan when it comes out. But, it will push the little GPU miners out like myself. You should just leave the poll up for a bit and see what the people think.

[sammy007]

Most mining pools do not support Scrypt N at the moment. Would we be doing more harm than good?

I am not sure 100%, but they don't care what algo is, because they need just a daemon. Am I missing something?

[DigiByte]

Most mining pools do not support Scrypt N at the moment. Would we be doing more harm than good?

I am not sure 100%, but they don't care what algo is, because they need just a daemon. Am I missing something?

We have never ran a pool but will look more into it.

[flounderella]

In the long run (think few years out), the network really can't run on gpus... if Digibyte becomes a serious online medium for exchanging goods and services, ASICs will play an important in securing the blockchain and maintaining a high hash rate at a fraction of the power. Even today 10 gridseeds can put out 3.5Mh of hashing power at just under 70 watts. GPUs waste energy. Just my two cents.

[Unpropitious]

If you decide to switch algorithm, please switch to something that does something else than forces the hashing power down. For example the vert mining just forces the hash down but also seems to cause a lot of strain on the gpu even at low intensity, which I find stupid. The best would be to implement something new if you decide to change, even though I do not really understand the need to change. However to have something unique might interest new investors also it would show that you are not just another coin dev that copy pastes code.

[DigiByte]

If you decide to switch algorithm, please switch to something that does something else than forces the hashing power down. For example the vert mining just forces the hash down but also seems to cause a lot of strain on the gpu even at low intensity, which I find stupid. The best would be to implement something new if you decide to change, even though I do not really understand the need to change. However to have something unique might interest new investors also it would show that you are not just another coin dev that copy pastes code.

You bring up some very valid points. We are looking more into each algorithm to decide if we can improve the performance and maybe implement something unique to DigiByte like we did with DigiShield.

[creativecuriosity]

In the long run (think few years out), the network really can't run on gpus... if Digibyte becomes a serious online medium for exchanging goods and services, ASICs will play an important in securing the blockchain and maintaining a high hash rate at a fraction of the power. Even today 10 gridseeds can put out 3.5Mh of hashing power at just under 70 watts. GPUs waste energy. Just my two cents.

With my understanding, this is inaccurate.

The actual "hash-rate" is unimportant.  Ergo, "pushing down" the hashrate is not a concern.

The important thing to understand is that hash is relative.  If everyone is hashing at, for arguments sake, 50% of the previous hash - then your reward for hashing remains the same. 

I haven't looked into x11, but the important different between sCrypt and n-factor sCrypt is the requirement of memory during the hash function.  The CPU algorithms force the limiting factor to be memory, as opposed to CPU/GPU/ASIC.  Memory, it seems at least, is not poised to experience the exponential growth in price/power.  This is what makes it attractive as a limiting factor.

Instead of paying any attention at all to "hashing power", the conversation needs to be about DISTRIBUTED "hashing power".  A network could have the highest "hashing power" of any coin in existence and at the same time be the least secure chain.  Keeping the hashing power distributed and not centralized to a few actors is the ultimate, and only important, goal that should be considered.

[milkyone]

let the asics have digibyte and hold all your coins will only drive value up like bitcoin. saying that im only a one card wonder gpu farms have more to loose NEED MORE MERCHANTS ALL THESE COINS ITS A RACE AGAINT TIME WHO GET ESTABLISHED FIRST. the cyrpto coin grave yard is stacking up

[DigiByte]

let the asics have digibyte and hold all your coins will only drive value up like bitcoin. saying that im only a one card wonder gpu farms have more to loose NEED MORE MERCHANTS ALL THESE COINS ITS A RACE AGAINT TIME WHO GET ESTABLISHED FIRST. the cyrpto coin grave yard is stacking up

You are absolutely correct. It is a race to see who gets established first. We are doing everything we can to be that 1st coin. We are in this for the long haul!

[benjamoyne]

I think that an alternative algorithm should be developed (maybe in conjunction with the Doge coin developers??)  but held off until we see how hard the network is hit by the asics.
It may not be as big a problem as people are thinking.

We need to remember there are people out there running farms with hundreds of gpus. These are the same people who will invest big in the new asics, so yes the hash rate will grow, but by how much, as I'm sure these are the same people who will turn off there gpu farms and mine purely on asics.

Maybe work with the people behind Sgminer directly to develop the new algorithm, im sure between Digi, Doge and the Sgminer crew you could come up with the true next gen algorithm.

[DigiByte]

I think that an alternative algorithm should be developed (maybe in conjunction with the Doge coin developers??)  but held off until we see how hard the network is hit by the asics.
It may not be as big a problem as people are thinking.

We need to remember there are people out there running farms with hundreds of gpus. These are the same people who will invest big in the new asics, so yes the hash rate will grow, but by how much, as I'm sure these are the same people who will turn off there gpu farms and mine purely on asics.

Maybe work with the people behind Sgminer directly to develop the new algorithm, im sure between Digi, Doge and the Sgminer crew you could come up with the true next gen algorithm.

This is a sensible approach to take. We have had some discussions with the Doge devs about this. We feel at most there is going to only be a 3-5 fold performance increase. No where near like what happened with Bitcoin Asics.

There are a number of issues that will be experienced with making another hard fork. During the last one we lost several mining pools, an exchange and all our block explorers. As DigiByte grows and more companies/platforms add DigiByte it will make hard forking and switching algos less of a possibility.

[benjamoyne]

I think that an alternative algorithm should be developed (maybe in conjunction with the Doge coin developers??)  but held off until we see how hard the network is hit by the asics.
It may not be as big a problem as people are thinking.

We need to remember there are people out there running farms with hundreds of gpus. These are the same people who will invest big in the new asics, so yes the hash rate will grow, but by how much, as I'm sure these are the same people who will turn off there gpu farms and mine purely on asics.

Maybe work with the people behind Sgminer directly to develop the new algorithm, im sure between Digi, Doge and the Sgminer crew you could come up with the true next gen algorithm.

This is a sensible approach to take. We have had some discussions with the Doge devs about this. We feel at most there is going to only be a 3-5 fold performance increase. No where near like what happened with Bitcoin Asics.

There are a number of issues that will be experienced with making another hard fork. During the last one we lost several mining pools, an exchange and all our block explorers. As DigiByte grows and more companies/platforms add DigiByte it will make hard forking and switching algos less of a possibility.

Also given the very cheap pricing of some of the Asics (www.fibonacci.io has 3460khs for about 22 lite coins) Normal miners can afford them, hell this one works out cheaper than gpus!

[bigc1984]

What sort of hash rates to you see with X11 & Scrypt-N is it about the same as Scrypt? Or is their a reduction in hash output?

Never bothered with Scrypy-N, But with x11 I get the following:

1.8 mh/s on an Asus 7970
1.24 mh/s on a R9-270

So you are actually getting a hash increase over Scrypt/CGminer?

Yep. Different Algorithms so different hashrates

         Scrypt Equiv    x11
7970    720kh/s         1.8 mh/s
270     480kh/s          1.24 mh/s

Interesting, thanks for sharing.

Was talking to the Doge devs and they brought up some good points. The stratum on several pools will have to be changed and at the exact block of change we could see some wacky things occur with the net hash rate.

Most mining pools do not support Scrypt N at the moment. Would we be doing more harm than good?

X11 is superior to scrypt N. I actually hate scrypt N =/

[maranello1561]

No way.
ASICs are the natural evolution to cheaply secure the blockchain... unless of course you want to preserve the massive gpu rigs of the big fish. For most small miners, ASICs will be much cheaper to buy and run... and maybe y'all can go back to playing Call of Duty with your GPUs instead of fucking up all the gamers by driving prices of gpus to unreasonable levels

[HollowStorm]

I vote for useful POW. Something like GridCoin, RieCoin. Something that will use hashing power noy only to secure the network but also for some really useful computing!

Does anyone have more examples like Grid, Rie, maybe? Has anyone heard of curecoin?

I believe the Curecoin devs have an algo ready but they didnt lauch the coin. Can we collaborate with them?

Check this out:
"CureCoin allows owners of both ASIC and GPU/CPU hardware to earn. CureCoin puts ASICs to work at what they are good at--securing a blockchain, while it puts GPUs and CPUs to work with work items that can only be done on them--protein folding. "

https://bitcointalk.org/index.php?topic=330685.0

Everyone Wins!

[haggis]

I read about a coin these days with multiple algorithms - each with its own difficulty and block times - but with the same coins per timeframe at the end.
This sounds very promising to me as everyone can mine the algo he/she deserves without harming gpu, cpu or asic users. They all have the same chance to find a block.

However, I can't remember the name of that coin