The creator of Ethereum, Vitalik Buterin, talked about his vision for enabling global open platforms and positively impacting human economic coordination.
This will be via
* censorship resistance
* trust minimization
* Fraud Resistance
Vitalik discussed those goals and benefits in more detail after talking about Ethereum being vastly less efficient than regular computing.
Vitalik Buterin walked people through calculations of how Ethereum is one million times less efficient than Amazon EC2 for computing, network and storage at this time. Blockchains are NOT for simple raw efficiency.
Ethereum World Computer costs 1.4 million times more than Amazon EC2 Cloud for computation/h2>
Looking at Amazon EC2 (cloud) pricing. The computing cost is currently about 4 cents per hour.
The cost an Ethereum block is about $13.40. An Ethereum block is processed in 200 milliseconds. An hour of Ethereum computing is 18,000 times more.
One hour of Ethereum computing is 1.4 million times more expensive than Amazon EC2.
The Ethereum Network is 1 million times less efficient
The Etherum network size 15,000 nodes. There is a safety factor so the block interval is 14 seconds between blocks divided by the block verification time of 200 milliseconds. A node in the Etherium network spends about 1/70th of its time performing computation.
15,000 nodes each repeat the calculation. 70 times 15,000 nodes is 1.05 million.
Storage cost 1 million times more on Ethereum
Storage costs about $80 for 250 gigabytes harddrive. The Etherium world computer can fill one byte of contracts data for a cost of 200 gas. Assuming ether is at $420 then it costs $84 million dollars for 250 gigabytes. Storage costs are over 1 million times more.
Storage, network and computation all currently are about 1 million times more costly.
The Ethereum world computer is 1 million times less efficient than regular computers. Raw execution efficiency is not what blockchains give you.
Censorship Resistance goal and benefit
Having a distributed system enables blockchain applications to be resistant to governments, internet companies and banks. These kinds of censorship have been seen many times.
There has been censorship of the internet by Facebook and Twitter.
This is where the person, company or organization is trying to steal from you. This famously happened with the loss of crypto assets from Mt Gox. However, Mt Gox was a centralized exchange.
A decentralized system can be more robust through redundancy and not having single points of failure.
Blockchains can be made interoperable between different blockchains.
Many companies built their businesses on top of Facebook or Youtube. Facebook and Google have discontinued APIs or changed algorithms that wiped out many companies and businesses.
Currently only public blockchains are willing to accept the 1 million times lower efficiency
Currently only cryptocurrency speculation and Silk road are the only things willing to accept the 1 million times lower efficiency of blockchain on a sustained basis.
We will not be able to get to helping the developing world with identification and microtransactions or many other applications until there is more scalability.
Sharding Plasma Channels and other technologies could make Cryptocurrencies like Ethereum 1000 times more efficient
Sharding is where every node did not have to process every operation. The network is divided into sections which operate semi-independently.
Vitalik’s Ethereum Research group is developing called shards. Sharding is called a Layer 1 scaling solution because it is implemented at the base-level protocol of Ethereum.
Layer 2 solutions like state channels, Plasma, and Truebit can provide large efficiency gains. They enable more transactions from the same capacity.
Plasma is composed of two key parts of the design:
* Reframing all blockchain computation into a set of MapReduce functions
* an optional method to do Proof-of-Stake token bonding on top of existing blockchains with the understanding that the Nakamoto Consensus incentives discourage block withholding.
Developers are willing to code in Python versus C despite Python being about 100 times lower execution speed. They trade off speed for simplicity and increased development speed.
Only the business logic that absolutely needs consensus needs to have the 1000X penalty. The rest can be built in normal coding.
Public Blockchains also have issues with Latency
It can take 33 minutes to get final confirmation of a blockchain transaction. 14 seconds for first confirmation of an Ethereum transaction.
There are different kinds of security risks.
Layer 2 solutions like state channels, Plasma, and Truebit can move more activity off chain. They enable more transactions from the same capacity and can reduce latency. They can also enhance privacy.
Vitalik gives the example of the Plasma Exit game. The game is implemented and setup on chain but there is off change messaging which can affect the game. Messages are eventually sent for resolution on chain.
This increases the cost of the security.
State channels can solve the latency issues. Signed messages are sufficient to authorize withdrawals without confirmations. The user has finalty once the signed message is received.
There are also designs where money cannot be stolen via hacking. However, hacking could force a temporary shutdown of the system to resolve the transactions in a solution which did not allow the stealing of assets.
The Future of Ethereum and blockchain will be about building applications which properly tradeoff more efficient solutions with other desired features and goals.