|
joshki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 03, 2017, 04:36:28 PM |
|
The general concept of proof of work is that you need a lot of hash power to secure the network so no one can build a second chain and double spend.
It's hard to do this with gpus. It's easier with custom hardware.
That's all.
|
|
|
|
margana (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 03, 2017, 05:04:18 PM |
|
The general concept of proof of work is that you need a lot of hash power to secure the network so no one can build a second chain and double spend.
It's hard to do this with gpus. It's easier with custom hardware.
That's all.
I can't really follow that logic. What is the difference between having 1,000 GPUs or 100 ASIC or whatever the ratio might be? On the contrary from that perspective I would come to the opposite conclusion. It would seem like ASIC ("application specific integrated circuits") would lend themselves to more risk of centralization and concentration of the total network power in a few select hands. There are less designers and manufacturers of ASIC chips appropriate for mining any given crypto currency (a very specific piece of hardware) as compared to something as general as GPUs. If you have the money to invest you can almost always find a bunch of GPUs to go out and buy, but that isn't the case unless you have the resources to design and build your own ASIC chips. You have to put down a bunch of money far in advance and hope that your order will get delivered on time (or close enough to on time) that the hardware you bought is still going to be relevant. Seems to me like a currency can be mined on GPUs but doesn't lend itself well to ASICs (or doesn't have ASICs yet available) is more secure as it is much easier for a wider variety of distributed players to participate in the network. What am I missing?
|
|
|
|
joshki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 01:50:06 AM |
|
It's hard to buy a thousand GPUs.
Yes, ASICs do contribute somewhat to centralization -- that is an issue. But look at the hashrate for bitcoin compared to some of the altcoins with asic-hostile hash algorithms, and compare the difficulty of a double-spend on the btc network compared to how little hardware it would take to overwhelm a network like sia or one of those.
You'd have to spend billions to double spend on btc. It might take a few million to overwhelm and take over the network on some of the altcoins and build your own chain.
|
|
|
|
philipma1957
Legendary
Offline
Activity: 4242
Merit: 8478
'The right to privacy matters'
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 02:13:04 AM |
|
It is a good debate. A large asic proof coin such as Zec is very secure. But little hash rate coins are not.
It would be costly to kill ZEC
IT WOULD BE MORE COSTLY TO KILL BTC
|
|
|
|
Khurram Bin Kamal
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 02:22:32 AM |
|
First you have to understand the network security or 51% attack The more network hashrate it is secure more So now if any one wants to attack your blockchain and destroy it or ruined it if that network is based on GPU only he just have to point his GPU towards your blockchain destroy it and go again mining some other coins and there are some pools powerful enough to destroy many altcoins And he loses nothing except for a certain time of mining lost and his GPU will go mine other altcoins
Now come to ASIC miners If a network is based on asic miners the attacker has to build or purchase that specific algorithm miner and that costs a lot and for every algorithm there is only 1 major coin with some minor coins and if he some how manages to get hold of 51% mining power of asics which is very hard to get and very capital intensive if he destroys that blockchain than his miners will also be of no use and he will loose a substantial amount also So there is a huge monetary loss to him also which no one can afford That's why blockchains with asic are more secure than those who are GPU based
|
|
|
|
philipma1957
Legendary
Offline
Activity: 4242
Merit: 8478
'The right to privacy matters'
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 03:56:38 AM |
|
First you have to understand the network security or 51% attack The more network hashrate it is secure more So now if any one wants to attack your blockchain and destroy it or ruined it if that network is based on GPU only he just have to point his GPU towards your blockchain destroy it and go again mining some other coins and there are some pools powerful enough to destroy many altcoins And he loses nothing except for a certain time of mining lost and his GPU will go mine other altcoins
Now come to ASIC miners If a network is based on asic miners the attacker has to build or purchase that specific algorithm miner and that costs a lot and for every algorithm there is only 1 major coin with some minor coins and if he some how manages to get hold of 51% mining power of asics which is very hard to get and very capital intensive if he destroys that blockchain than his miners will also be of no use and he will loose a substantial amount also So there is a huge monetary loss to him also which no one can afford That's why blockchains with asic are more secure than those who are GPU based
Close but wrong. I am bitmain tons of cheap gear and cheap power. I destroy btc. And switch to bcc my mining gear is good. Which is the point of the op. My point is if bitmain did this and succeeded it killing btc then switched to bcc the risk is bcc might fail. If you study the switching to bcc where it almost had more hash power then btc. I would say ASIC mining is not as secure as it was before bcc was created.
|
|
|
|
joshki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 04:53:08 AM |
|
First you have to understand the network security or 51% attack The more network hashrate it is secure more So now if any one wants to attack your blockchain and destroy it or ruined it if that network is based on GPU only he just have to point his GPU towards your blockchain destroy it and go again mining some other coins and there are some pools powerful enough to destroy many altcoins And he loses nothing except for a certain time of mining lost and his GPU will go mine other altcoins
Now come to ASIC miners If a network is based on asic miners the attacker has to build or purchase that specific algorithm miner and that costs a lot and for every algorithm there is only 1 major coin with some minor coins and if he some how manages to get hold of 51% mining power of asics which is very hard to get and very capital intensive if he destroys that blockchain than his miners will also be of no use and he will loose a substantial amount also So there is a huge monetary loss to him also which no one can afford That's why blockchains with asic are more secure than those who are GPU based
Close but wrong. I am bitmain tons of cheap gear and cheap power. I destroy btc. And switch to bcc my mining gear is good. Which is the point of the op. My point is if bitmain did this and succeeded it killing btc then switched to bcc the risk is bcc might fail. If you study the switching to bcc where it almost had more hash power then btc. I would say ASIC mining is not as secure as it was before bcc was created. Yet btc continues.
|
|
|
|
Khurram Bin Kamal
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 07:13:01 AM |
|
First you have to understand the network security or 51% attack The more network hashrate it is secure more So now if any one wants to attack your blockchain and destroy it or ruined it if that network is based on GPU only he just have to point his GPU towards your blockchain destroy it and go again mining some other coins and there are some pools powerful enough to destroy many altcoins And he loses nothing except for a certain time of mining lost and his GPU will go mine other altcoins
Now come to ASIC miners If a network is based on asic miners the attacker has to build or purchase that specific algorithm miner and that costs a lot and for every algorithm there is only 1 major coin with some minor coins and if he some how manages to get hold of 51% mining power of asics which is very hard to get and very capital intensive if he destroys that blockchain than his miners will also be of no use and he will loose a substantial amount also So there is a huge monetary loss to him also which no one can afford That's why blockchains with asic are more secure than those who are GPU based
Close but wrong. I am bitmain tons of cheap gear and cheap power. I destroy btc. And switch to bcc my mining gear is good. Which is the point of the op. My point is if bitmain did this and succeeded it killing btc then switched to bcc the risk is bcc might fail. If you study the switching to bcc where it almost had more hash power then btc. I would say ASIC mining is not as secure as it was before bcc was created. How can one be close and wrong simultaneously If you were bitmain and you can't destroy btc with 16 % hashrate and even you manage to get 51% hashrate you will not destroy btc because the faith of the world is in btc not in Bch and if btc gets killed Bch will go with it and your machines will be of no use and also the future sales of machines will be affected and your company will be closed So that's why ASIC are the security of a BLOCKCHAIN NETWORK
|
|
|
|
margana (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 07:55:44 AM |
|
I can see the points of both sides, however I remain unconvinced entirely by either side, however intuitively I feel like GPU mining in the hands of many user/miners is better than large mining farms with expensive specialized hardware.
Let's do a thought experiment and make some assumptions.
Let's say cyrptocurrency A is GPU minable and cryptocurrency B is ASIC mineable.
ASIC of reasonable power for mining for cryptocurrency B cost $1,000 USD each. You have a few miner players but mostly large farms with thousands of machines in each. Let's say there are 100,000 of these machines with 20% spread across smaller players and the remaining 80% spread across 5 large players.
GPUs with reasonable mining power for cryptocurrency A cost $100 USD each. You have 800,000 user/miners running 1 million cards. Most of these people just have one or two graphics card running. You have just a flew players with large GPU mining farms.
It seems to me like you can make a good argument that cryptocurrency A is more secure, at least from the point of being resistant to centralized control and subversion of the system. And isn't that whole point with cyrptocurrencies? To avoid centralization?
Let me put this another way, we can remove ASIC from the discussion and just talk about hash power. Let's say network A has 1 gigahash of power and so does network B. However, network A has 100,000 machines capable of 10,000 hashes per second distributed between 10,000 different miners. Network B has 1,000,000 machines capable of 1,000 hashes per second each distributed between 800,000 miners. Wouldn't network B be preferable to network A in terms of security, decentralization and resistenance to manipulation in the direction of the concurrency by small subset(s) of the network participants?
As as side note, aren't there a lot of things you can mine with a bitcoin ASIC? I know even as of a few years ago there were a few different things you could do with it. Does anyone have a definitive list of all the cryptocurrencies you can also mine with a bitcoin ASIC?
|
|
|
|
joshki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 07:59:30 AM |
|
I can see the points of both sides, however I remain unconvinced entirely by either side, however intuitively I feel like GPU mining in the hands of many user/miners is better than large mining farms with expensive specialized hardware.
Let's do a thought experiment and make some assumptions.
Let's say cyrptocurrency A is GPU minable and cryptocurrency B is ASIC mineable.
ASIC of reasonable power for mining for cryptocurrency B cost $1,000 USD each. You have a few miner players but mostly large farms with thousands of machines in each. Let's say there are 100,000 of these machines with 20% spread across smaller players and the remaining 80% spread across 5 large players.
GPUs with reasonable mining power for cryptocurrency A cost $100 USD each. You have 800,000 user/miners running 1 million cards. Most of these people just have one or two graphics card running. You have just a flew players with large GPU mining farms.
It seems to me like you can make a good argument that cryptocurrency A is more secure, at least from the point of being resistant to centralized control and subversion of the system. And isn't that whole point with cyrptocurrencies? To avoid centralization?
As as side note, aren't there a lot of things you can mine with a bitcoin ASIC? I know even as of a few years ago there were a few different things you could do with it. Does anyone have a definitive list of all the cryptocurrencies you can also mine with a bitcoin ASIC?
any of the sha256 btc clones are mineable with a btc asic. A GPU can mine anything that's had the code built for it.
|
|
|
|
margana (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 08:01:10 AM |
|
any of the sha256 btc clones are mineable with a btc asic. A GPU can mine anything that's had the code built for it.
Where is there a definitive list of which those are?
|
|
|
|
joshki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 08:02:13 AM |
|
any of the sha256 btc clones are mineable with a btc asic. A GPU can mine anything that's had the code built for it.
Where is there a definitive list of which those are? any of the coin list sites. There are new ones made every day, most of them never go anywhere.
|
|
|
|
margana (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 02:12:56 PM |
|
any of the coin list sites. There are new ones made every day, most of them never go anywhere.
Such as which one for example? Coinmarketcap.com seems to have a very compreshenskce list but I don't see a column that indicates the algorithm?
|
|
|
|
joshki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 02:17:12 PM |
|
any of the coin list sites. There are new ones made every day, most of them never go anywhere.
Such as which one for example? Coinmarketcap.com seems to have a very compreshenskce list but I don't see a column that indicates the algorithm? https://www.cryptocompare.com/coins/#/btc
|
|
|
|
margana (OP)
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
September 04, 2017, 09:34:22 PM |
|
any of the coin list sites. There are new ones made every day, most of them never go anywhere.
Such as which one for example? Coinmarketcap.com seems to have a very compreshenskce list but I don't see a column that indicates the algorithm? https://www.cryptocompare.com/coins/#/btcThanks that's perfect! Does SHA256D work with the same ASICs as bitcoin's SHA256 or does it need a different type of chip?
|
|
|
|
|