Core needs to prepare a GPU only PoW - Spondoolies CEO Guy Corem - page 2.

iCEBREAKER

legendary

Activity: 2156

Merit: 1072

Crypto is the separation of Power and State.

Quote from: -ck on January 25, 2016, 10:52:33 PM

Quote from: iCEBREAKER on January 25, 2016, 10:28:49 PM

Quote from: -ck on January 25, 2016, 04:18:48 PM

The increasing number of people smoking some bad juju and coming up with this pipe-dream shit astounds me.

Are you talking about XT/Classic style contentious hard forks, having a backup plan to change the PoW if necessary, both, or neither?

Change PoW.

Contentious is not the same as absurd.

Note this is not a vote for anything in particular and should not be taken as such.

IMO, changing the PoW is not something to be done except as a last resort to protect the blockchain from a ~51% attacker.

But in a doomsday scenario, with an ongoing contentious hard fork causing unacceptable risk of catastrophic consensus failure, I'd appreciate having some good 2nd strike options having been presciently filed away on the shelf for future consideration in exigent circumstances.

It's like having a fire hose behind glass. Not a good idea to play with most of the time, but if there's a fire then it's time to break the glass and deploy emergency countermeasures.

In your opinion, is Bitcoin married to SHA-256 like it's married to other control variables like the 21e6 coin emission schedule, 10 minute block target, and 1MB blocks?

Is there a better way to resolve a (messy, scary, dangerous, mutually destructive) fork war than by one of the chains moving to another PoW?

I don't see Bitcoin as being defined by SHA-256; IMO the particular PoW (while carefully chosen as the best fit for the job) isn't an intrinsic feature of the Nakamoto Consensus protocol.

Is there something special about SHA-256 that other functions cannot do as well or better?

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Quote from: xyzzy099 on January 25, 2016, 11:18:03 PM

I understand what you are saying... I'm just trying to clarify that the price is not defined only by the cost of mining. The cost of mining sets a floor for the price, and demand can potentially pull the price to any level at all above that floor - and the price of mining will be pulled up at the same time, as it will be profitable (and ultimately necessary) for miners to spend more money to mine, no matter what the mining mechanism is (CPU, GPU, ASIC, whatever).

Usually the serious demand always first go to mining, since that is the lowest cost and will give them freshly mined coins. Only in case there is a sure anticipation of huge jump in difficulty due to new mining technology, they will go to exchange

If you have so serious problem inside bitcoin that you even need to change PoW, the chance that the demand would still rise around the fork period is almost 0

xyzzy099

legendary

Activity: 1066

Merit: 1098

Quote from: johnyj on January 25, 2016, 11:09:39 PM

Quote from: xyzzy099 on January 25, 2016, 10:57:01 PM

Quote from: johnyj on January 25, 2016, 10:48:48 PM

Quote from: xyzzy099 on January 25, 2016, 03:04:22 PM

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

I think you have it backwards. The cost of mining does not define the value of the coin. The amount of money people will be willing to pour into mining the coin is, however, determined by the value of the coin.

Do you know arbitrage? This has been discussed for years in the economy forum, I thought it is already a basic knowledge, just search for it there

Put it simply, if it cost $1 to mine one bitcoin and the market price is $400, everyone will mine and immediately sell, no one will buy; and those who want to hold coin will immediately sell their coins and mine it back at a much lower cost. That is $399 profit for every $1 spent, 399x return on investment. No matter how large the market demand is, the arbitraging will make sure the price keep dropping until it reaches the current mining cost

So, unless the mining cost quickly rise up to $400 overnight, a new pow coin with neglectable initial mining cost will just crash to zero in a very short time

You are assuming no real demand for bitcoin except speculation, which I expect will not be the case forever. If there is real demand, the demand will drive the price, and the cost of mining will rise to the value set by demand - not the other way around. If, on the other hand, demand remains purely speculative, then you would probably be correct that the price will fall to the floor established by the mining cost.

Of course there is market demand, but if I need one bitcoin, and it takes $1 to mine but $400 to buy, should I mine or buy? Maybe I don't have the necessary tools to mine, but I'm quite sure there will be plenty of miners who are willing to sell to me at $10, which already bring them a 1000% profit immediately

Maybe there are many people who are so stupid that they blindly buy at $400 a coin because they don't know the PoW has changed. But if there is only one guy knows this, he will immediately borrow as much coin as possible and at the same time buy as much hash power as possible, and dump the coin on exchange and mine them back, and in a few days everyone knows how to get rich quick

I understand what you are saying... I'm just trying to clarify that the price is not defined only by the cost of mining. The cost of mining sets a floor for the price, and demand can potentially pull the price to any level at all above that floor - and the price of mining will be pulled up at the same time, as it will be profitable (and ultimately necessary) for miners to spend more money to mine, no matter what the mining mechanism is (CPU, GPU, ASIC, whatever).

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Quote from: xyzzy099 on January 25, 2016, 10:57:01 PM

Quote from: johnyj on January 25, 2016, 10:48:48 PM

Quote from: xyzzy099 on January 25, 2016, 03:04:22 PM

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

I think you have it backwards. The cost of mining does not define the value of the coin. The amount of money people will be willing to pour into mining the coin is, however, determined by the value of the coin.

Do you know arbitrage? This has been discussed for years in the economy forum, I thought it is already a basic knowledge, just search for it there

Put it simply, if it cost $1 to mine one bitcoin and the market price is $400, everyone will mine and immediately sell, no one will buy; and those who want to hold coin will immediately sell their coins and mine it back at a much lower cost. That is $399 profit for every $1 spent, 399x return on investment. No matter how large the market demand is, the arbitraging will make sure the price keep dropping until it reaches the current mining cost

So, unless the mining cost quickly rise up to $400 overnight, a new pow coin with neglectable initial mining cost will just crash to zero in a very short time

You are assuming no real demand for bitcoin except speculation, which I expect will not be the case forever. If there is real demand, the demand will drive the price, and the cost of mining will rise to the value set by demand - not the other way around. If, on the other hand, demand remains purely speculative, then you would probably be correct that the price will fall to the floor established by the mining cost.

Of course there is market demand, but if I need one bitcoin, and it takes $1 to mine but $400 to buy, should I mine or buy? Maybe I don't have the necessary tools to mine, but I'm quite sure there will be plenty of miners who are willing to sell to me at $10, which already bring them a 1000% profit immediately

Maybe there are many people who are so stupid that they blindly buy at $400 a coin because they don't know the PoW has changed. But if there is only one guy knows this, he will immediately borrow as much coin as possible and at the same time buy as much hash power as possible, and dump the coin on exchange and mine them back, and in a few days everyone knows how to get rich quick

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Quote from: Carlton Banks on January 25, 2016, 03:07:47 PM

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

Is that your case for "why bitcoin will appreciate forever", Jonny? In any case, your assertion doesn't withstand a basic reality test: if it were true, it's conceivable that mining costs would be way, way higher than at present. Seeing as it would provoke a 3rd (4th?) mining gold rush.

That will be the case, the mining cost should always be close to bitcoin's market price, otherwise arbitraging will happen to make them equal

Due to that the block reward are less and less, you might be able to maintain the same cost while have price double every 4 years, in case block space are always plenty thus the fee income is always a small part of the block reward

xyzzy099

legendary

Activity: 1066

Merit: 1098

Quote from: johnyj on January 25, 2016, 10:48:48 PM

Quote from: xyzzy099 on January 25, 2016, 03:04:22 PM

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

I think you have it backwards. The cost of mining does not define the value of the coin. The amount of money people will be willing to pour into mining the coin is, however, determined by the value of the coin.

Do you know arbitrage? This has been discussed for years in the economy forum, I thought it is already a basic knowledge, just search for it there

Put it simply, if it cost $1 to mine one bitcoin and the market price is $400, everyone will mine and immediately sell, no one will buy; and those who want to hold coin will immediately sell their coins and mine it back at a much lower cost. That is $399 profit for every $1 spent, 399x return on investment. No matter how large the market demand is, the arbitraging will make sure the price keep dropping until it reaches the current mining cost

So, unless the mining cost quickly rise up to $400 overnight, a new pow coin with neglectable initial mining cost will just crash to zero in a very short time

You are assuming no real demand for bitcoin except speculation, which I expect will not be the case forever. If there is real demand, the demand will drive the price, and the cost of mining will rise to the value set by demand - not the other way around. If, on the other hand, demand remains purely speculative, then you would probably be correct that the price will fall to the floor established by the mining cost.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: iCEBREAKER on January 25, 2016, 10:28:49 PM

Quote from: -ck on January 25, 2016, 04:18:48 PM

The increasing number of people smoking some bad juju and coming up with this pipe-dream shit astounds me.

Are you talking about XT/Classic style contentious hard forks, having a backup plan to change the PoW if necessary, both, or neither?

Change PoW.

Contentious is not the same as absurd.

Note this is not a vote for anything in particular and should not be taken as such.

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Quote from: xyzzy099 on January 25, 2016, 03:04:22 PM

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

I think you have it backwards. The cost of mining does not define the value of the coin. The amount of money people will be willing to pour into mining the coin is, however, determined by the value of the coin.

Do you know arbitraging? This has been discussed for years in the economy forum, I thought it is already a basic knowledge, just search for it there

Put it simply, if it cost $1 to mine one bitcoin and the market price is $400, everyone will mine and immediately sell, no one will buy; and those who want to hold coin will immediately sell their coins and mine it back at a much lower cost. That is $399 profit for every $1 spent, 399x return on investment. No matter how large the market demand is, the arbitraging will make sure the price keep dropping until it reaches the current mining cost

So, unless the mining cost quickly rise up to $400 overnight, a new pow coin with neglectable initial mining cost will just crash to zero in a very short time

iCEBREAKER

legendary

Activity: 2156

Merit: 1072

Crypto is the separation of Power and State.

Quote from: -ck on January 25, 2016, 04:18:48 PM

The increasing number of people smoking some bad juju and coming up with this pipe-dream shit astounds me.

Are you talking about XT/Classic style contentious hard forks, having a backup plan to change the PoW if necessary, both, or neither?

Dogedarkdev

legendary

Activity: 1708

Merit: 1009

$XVG - The Standard in Crypto as a Currency!

Quote from: Dogedarkdev on January 25, 2016, 04:16:03 PM

Quote from: iCEBREAKER on January 25, 2016, 01:34:21 AM

https://medium.com/@vcorem/lesson-learned-from-the-classic-coup-attempt-or-why-core-needs-to-prepare-a-gpu-only-pow-6a9afe18e4b0

So, as I write these words, it seems that Classic has a good chance of meeting its activation Threshold. Jonathan is aiming for Classic to be activated when 75% of the last 2,016 (currently coded 1,000 but he said in the interview that he might increase to 2,016) mined blocks indicate support of the Classic fork, meaning the miners that mined them are ready to do a 2MB Hard Fork and leave behind everyone on Core.
What will happen to Core chain if Classic will be activated?

https://coin.dance/nodes

there's hardly any nodes running classic.. what do you mean it has a good chance?

oh i see now, there are no intentions in Classic to change the algo, it is strictly about block size increase from 1mb to 2mb.. OP was misleading..

Erkallys

legendary

Activity: 1120

Merit: 1004

Quote from: Dogedarkdev on January 25, 2016, 04:16:03 PM

Quote from: iCEBREAKER on January 25, 2016, 01:34:21 AM

https://medium.com/@vcorem/lesson-learned-from-the-classic-coup-attempt-or-why-core-needs-to-prepare-a-gpu-only-pow-6a9afe18e4b0

So, as I write these words, it seems that Classic has a good chance of meeting its activation Threshold. Jonathan is aiming for Classic to be activated when 75% of the last 2,016 (currently coded 1,000 but he said in the interview that he might increase to 2,016) mined blocks indicate support of the Classic fork, meaning the miners that mined them are ready to do a 2MB Hard Fork and leave behind everyone on Core.
What will happen to Core chain if Classic will be activated?

https://coin.dance/nodes

there's hardly any nodes running classic.. what do you mean it has a good chance?

This is a nice website you submitted us. I think that it should be the nodes + the miners that should be able to decide to fork or not, not only the miners.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

The increasing number of people smoking some bad juju and coming up with this pipe-dream shit astounds me.

Dogedarkdev

legendary

Activity: 1708

Merit: 1009

$XVG - The Standard in Crypto as a Currency!

Quote from: iCEBREAKER on January 25, 2016, 01:34:21 AM

https://medium.com/@vcorem/lesson-learned-from-the-classic-coup-attempt-or-why-core-needs-to-prepare-a-gpu-only-pow-6a9afe18e4b0

So, as I write these words, it seems that Classic has a good chance of meeting its activation Threshold. Jonathan is aiming for Classic to be activated when 75% of the last 2,016 (currently coded 1,000 but he said in the interview that he might increase to 2,016) mined blocks indicate support of the Classic fork, meaning the miners that mined them are ready to do a 2MB Hard Fork and leave behind everyone on Core.
What will happen to Core chain if Classic will be activated?

https://coin.dance/nodes

there's hardly any nodes running classic.. what do you mean it has a good chance?

Carlton Banks

legendary

Activity: 3430

Merit: 3080

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

Is that your case for "why bitcoin will appreciate forever", Jonny? In any case, your assertion doesn't withstand a basic reality test: if it were true, it's conceivable that mining costs would be way, way higher than at present. Seeing as it would provoke a 3rd (4th?) mining gold rush.

xyzzy099

legendary

Activity: 1066

Merit: 1098

Quote from: johnyj on January 25, 2016, 02:56:02 PM

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

I think you have it backwards. The cost of mining does not define the value of the coin. The amount of money people will be willing to pour into mining the coin is, however, determined by the value of the coin.

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Mining cost is the baseline of coin value, if you change to GPU mining, coins value will drop to single digits thus abandoned by investors, better use litecoin in that case

plorph

member

Activity: 76

Merit: 14

Quote from: nichu on January 25, 2016, 05:34:10 AM

wonderful changes ,looking forward for that, i liked to concept of GPU mining ,which everyone can participate rather than big mongers who owns huge number of ASICS Wink

I don't understand. Whatever the algorithms are, there will always be specialized chips made with enough time. If a change allows regular GPU mining for now, I'm sure it won't be long until specialized GPU's that maximize mining efficiency will come out.

Carlton Banks

legendary

Activity: 3430

Merit: 3080

Quote from: iCEBREAKER on January 25, 2016, 07:55:32 AM

I'm not sold on super-complex functions with Byzantine rules (despite the fact when I tried to write one, it looked a lot like what Guy came up with)! Cheesy

More complexity just raises the barrier to entry for ASIC manufacturers, exacerbating the centralization we already see with fairly simple SHA-256. Not that working at 16nm is simple...but still, buying proprietary IP for up to all 100 isn't going to be cheap! Sad

And even with ASIC designs for all 100 algorithms, they still can't outcompete thousands (or possibly even hundreds of thousands next wave) of GPUs and FPGAs, and certainly cannot be as quickly up-and-hashing. And what then happens when 100 becomes 101? Cheesy

Quote from: iCEBREAKER on January 25, 2016, 07:55:32 AM

Since we can't defeat all ASICs, perhaps we should try to encourage as much competition as possible and KISS.

Not convinced about that logic when current ASIC overlords are trying to act like they have a gun to Core's head, but we'll see to what extent the credibility of their threat unravels soon enough.

iCEBREAKER

legendary

Activity: 2156

Merit: 1072

Crypto is the separation of Power and State.

Quote from: Carlton Banks on January 25, 2016, 06:39:48 AM

10 stacked hash functions, randomized from a set of 100? Rotated every 3 months? Lol plan delivers

Quote from: nichu on January 25, 2016, 05:34:10 AM

wonderful changes ,looking forward for that, i liked to concept of GPU mining ,which everyone can participate rather than big mongers who owns huge number of ASICS Wink

And the most well motivated to come up with algorithmic shortcuts (==moreMiningBTC) for the latest hash-stack monster will always be the most deserving; young, bright, hungry and
with loads of time to spare (day-to-day and long term). The Blockchain "Alliance" can stick that in their respective crackpipes.

I'm not sold on super-complex functions with Byzantine rules (despite the fact when I tried to write one, it looked a lot like what Guy came up with)! Cheesy

More complexity just raises the barrier to entry for ASIC manufacturers, exacerbating the centralization we already see with fairly simple SHA-256. Not that working at 16nm is simple...but still, buying proprietary IP for up to all 100 isn't going to be cheap! Sad

Since we can't defeat all ASICs, perhaps we should try to encourage as much competition as possible and KISS.

Carlton Banks

legendary

Activity: 3430

Merit: 3080

10 stacked hash functions, randomized from a set of 100? Rotated every 3 months? Lol plan delivers

Quote from: nichu on January 25, 2016, 05:34:10 AM

wonderful changes ,looking forward for that, i liked to concept of GPU mining ,which everyone can participate rather than big mongers who owns huge number of ASICS Wink

And the most well motivated to come up with algorithmic shortcuts (==moreMiningBTC) for the latest hash-stack monster will always be the most deserving; young, bright, hungry and
with loads of time to spare (day-to-day and long term). The Blockchain "Alliance" can stick that in their respective crackpipes.

Topic: Core needs to prepare a GPU only PoW - Spondoolies CEO Guy Corem - page 2. (Read 3446 times)