EDIT: a correction was made to the article to remove an erroneous claim that a Cardano account needs at least 1% of the total supply of ADA to participate in staking. I obtained this number from Dan Larimer's blog post on Steemit analysing Cardano. In actuality, 1% of total ADA supply is a suggested threshold from the original Ouroboros paper for accounts that want to participate in the random block producer selection process for a Cardano epoch. Although it is true that this will lead to centralization (by design, the algorithm necessitated a very small group of wealthy participants in the block producer selection process to maintain blockchain performance), according to Charles Hoskinson, a planned upgrade to Cardano's consensus algorithm will remove this threshold.
Cardano is a proof of stake, smart contract blockchain. Its native cryptocurrency is called ADA and the smallest unit of ADA (1/1,000,000 ADA) is called a Lovelace - in homage to the first recognized computer programmer, Ada Lovelace.
The Cardano team positions itself as one with a deeply academic background. The project’s website claims that Cardano is “the first blockchain project to be developed from a scientific philosophy, and the only one to be designed and built by a global team of leading academics and engineers”.
It’s apparent that Cardano has applied significant academic rigour in its development of its consensus algorithm, Ouroboros. Ouroboros is a provably secure proof of stake algorithm developed by a team of IOHK scientists led by Prof. Aggelos Kiayias. The algorithm is backed by a peer reviewed paper. According to Cardano founder Charles Hoskinson, the Ouroboros paper was rewritten 8 times before its publication.
The Cardano team has also published a set of other academic papers, including one for a scalable and decentralized random number generator algorithm, one for an efficient and decentralized poker protocol, and one for secure sidechains. But besides the random number generator algorithm, which is used in Ouroboros to select stake winners, Cardano has not incorporated ideas from the other papers (yet).
Unlike Ethereum, where the settlement layer and the computational layer are intertwined, Cardano is built as a layered blockchain.The Cardano Settlement Layer (SL) is where account balances and transactions are managed. It uses the UTXO model and is supposed to be the most secure part of the blockchain. The SL is kept apart from the computational layer where smart contracts are executed. Why build a layered blockchain? This layered architecture gives the system the flexibility to be more easily maintained and allow for upgrades by way of soft forks.
Cardano’s academic rigour and provable safety are unique in the cryptocurrency space, and should be appealing to enterprise customers. Companies looking to deploy blockchain technology want to do it in the most risk-free way possible and a cryptocurrency with a provably safe consensus algorithm backed by peer-reviewed scientific papers should be an enticing platform to deploy on.
Cardano’s ICO took place in 4 stages between September 2015 - January 2017 with strict Know Your Customer (KYC) requirements. It was primarily marketed to Japanese investors as an “investment to retire on.” In fact, 95% of the buyers were of Japanese origin, primarily in the 35-55 age bracket. 2.56% of buyers were Korean, and 2.39% were Chinese. The ICO oversaw a total sale of approximately 26,000,000,000 of the 45,000,000,000 maximum supply of ADA tokens. At the conclusion of the ICO a total of approximately $63 million USD was raised; which put the price at an average of $0.00242 per ADA, and the market capitalization immediately after the ICO at approximately $109 million USD (source).
While 26,000,000,000 ADA were sold in the ICO, an additional 5,185,414,108 ADA were distributed to IOHK, Emurgo, and the Cardano Foundation.
A total of 31,112,484,646 ADA was made available at Cardano’s launch.
Current state of Cardano
Cardano outlines three eras in its roadmap: the Testnet Era, Bootstrap Era, and Reward Era. The Testnet Era is pretty self-explanatory, Cardano is deployed on a test net.
The Bootstrap Era is when Cardano SL is deployed on the mainnet BUT it is centralized in order to maintain security. Only a fixed, predefined set of users have control over the system. The assumption is that security is poor in this era since people are slow to redeem their ADA coins. For the first few months, stakeholders who jointly possess the majority of the ADA stake are mostly not online yet. The blockchain can be easily attacked if only a small amount of ADA is staked. As consolation to the community, Ouroboros will not provide any block rewards.
Cardano is currently in the Bootstrap Era but is on the verge of transitioning to the Reward Era.
The Reward Era is when the network is in a completely decentralized and trustless mode. According to Cardano’s website, “the system will receive regular software updates moving forward, and a massive amount of new features will be released” in this era. These features would include upgrades to Ouroboros, Cardano’s consensus algorithm, an implementation of Cardano’s control layer, and sidechains.
IOHK and the Cardano Foundation
IOHK, or Input Output Hong Kong, is the engineering company responsible for implementing Cardano whereas the Cardano Foundation is the promoter, educator, and standard body for the blockchain and its apps. In October, Hoskinson stated that IOHK had nearly 100 employees and contractors, three research engineers, and operate in nearly ten countries.
Charles Hoskinson founded both IOHK and the Cardano Foundation. Hoskinson is a serial blockchain entrepreneur. In fact, he was initially the CEO of Ethereum before being fired. At the time, Ethereum at the time was 6 months old with a $15 million market cap. Reasons for the firing are murky.
Hoskinson went on to found IOHK, and IOHK subsequently led the development of Ethereum Classic (which came to be when Ethereum hard forked after the major DAO hack). A small portion of the community was against the fork and maintained the original chain, which is today known as Ethereum Classic.
While IOHK continues to lead Ethereum Classic’s development, its primary focus is now on Cardano. Hoskinson aspires for Cardano to be a compelling alternative to Ethereum, and he plans to do so by building a blockchain that is designed for scale, security, and upgradeability from day one.
Ouroboros is Cardano’s proof of stake consensus algorithm. A proof of stake consensus algorithm was chosen for Cardano since it’s more energy efficient and is more expensive to perform a 51% attack.
The name Ouroboros comes from the name of an ancient symbol depicting a serpent or dragon eating its own tail. The symbol originated in ancient Egyptian iconography. The name is fitting for this algorithm, since Ouroboros operates in epochs.
What does that mean? Each Ouroboros epoch contains a fixed number of slots. Each slot represents a new block. The slot leader is an account staking ADA that is responsible for creating the block. Epochs last for 5 days. Slot leaders for an epoch are chosen during the previous epoch. An account’s chance of being chosen is directly tied to the amount of ADA it’s staking.
According to the original Ouroboros paper, an account can only participate in an election committee that randomly selects block producers for each epoch if the ADA it holds and the votes it receives from the community exceeds a certain threshold. The paper's suggested threshold is 1% of total ADA supply. Any staking wallet with a non zero ADA balance can be selected to be a block producer. However, the election committee process is prone to centralization given the threshold requirement, which will greatly reduce the number of accounts that can participate. In fact, the original paper states that this election committee needs to be small in order to preserve blockchain performance. According to Charles Hoskinson, a planned upgrade to Ouroboros, Ouroboros Praos, will remove this threshold.
Ouroboros will initially provide both newly minted ADA and transaction fees as block rewards. When the total supply of ADA reaches 45,000,000,000 block rewards will only consist of transaction fees.
According to IOHK/Cardano Foundation, Ouroboros is the first provably secure proof of stake algorithm - meaning the team behind Cardano has mathematically proven that the algorithm is safe from all the known proof of stake attack methods including grinding attacks, nothing-at-stake attacks, and long range attacks. This safety and security serves as a very powerful enterprise selling point for Cardano.
Read more about Ouroboros here.
Ouroboros Praos is a planned upgrade to Ouroboros. Praos is semi-synchronous, doesn’t require a heavy MPC protocol for randomness, and encourages the adoption of a quantum resistant signature scheme for input endorsers and slot leaders. Praos allows for shorter block times. For a deeper dive into Praos, you can read the paper here.
Ouroboros Hydra is another upgrade to Ouroboros that introduces sharding. Sharding allows multiple Ouroboros epochs to run at the same time which significantly increases Cardano’s throughput. One of the hardest things to figure out for sharding is inter-shard communication.
We are still a ways away from the Hydra upgrade. Expect a paper from Cardano’s team on Hydra before any serious development efforts begin.
Sidechains are blockchains running in parallel and independently from the main chain. Sidechains and the main chain should be able to securely transfer assets back and forth. If any problems arise in the sidechain, the main chain can confiscate the sidechain’s assets.
Sidechains significantly extend the main chain’s capabilities. Because they are separate from the main chain, they can be more centralized, faster, and perform more specialized functions. By pegging themselves to the main chain, side chains can take advantage of the main chain’s security without having to adhere to the main chain’s inefficiencies (e.g. lower transaction throughput to optimize for decentralization, which increases security).
The main problem with sidechains today is their security. Specifically, there are several attacks where mainchains can be compromised through side chains. The Cardano team has attempted to solve this by introducing a provably secure sidechain technology that fixes all known avenues of attack, opening the way for secure sidechains. They call it Non-interactive Proofs of Proof of Work (NIPoPoW). Although the Cardano’s team is far from implementing NIPoPoW sidechains on Cardano, the technology is a significant step forward.
Cardano’s average transaction throughput is 10-15 transactions per second which is comparable to Ethereum. Charles Hoskinson has claimed that Ouroboros can perform at more than 200 transactions per second in a lab unsharded.
Smart contracts are not ready
Cardano’s smart contract virtual machine, the IELE virtual machine, is currently under development. According to Cardano’s roadmap, it is 75% complete.
Cardano’s smart contract language, Plutus, is also currently under development. The specification of Plutus is complete but the implementation, Plutus core, is not. According to the roadmap, it is 12% complete.
With the control layer still in the early stages of development, Cardano is currently a UTXO blockchain using a proof of stake consensus algorithm. In order for it to compete against existing decentralized smart contract blockchains like Ethereum and QTUM, Cardano needs to get out its control model ASAP in order to get its network effect going.
Ouroboros is prone to centralization
Ouroboros is prone to centralization until the release of Ouroboros Praos, a planned upgrade of the consensus algorithm. As mentioned above, an account can only participate in the random selection of block producers for an epoch if the ADA it holds and the votes it receives from the community exceeds a certain threshold. The original Ouroboros paper suggested a 1% of total ADA supply threshold.
This suggested minimum threshold requirement means that at most 100 accounts can participate in the block producer selection process and because of the power law, there will be far fewer than 100. There will most certainly be a few entities with exponentially large balances of ADA and the minimum threshold requirement will only serve to consolidate their power and control. The original paper states that by design, the number of accounts participating in the block producer selection process needs to be small to maintain blockchain performance.
The good news is, according to Charles Hoskinson, Ouroboros Praos will not have this minimum threshold requirement. This will increase the decentralization of Cardano.
Dan Larimer’s critique of Ouroboros
There is a long, contentious history between Dan Larimer, the founder of EOS, and Charles Hoskinson, the founder of Cardano. In fact, before Ethereum, Hoskinson actually worked with Larimer on Bitshares but left on bad terms.
Check out this post by Larimer on the flaws of Cardano's consensus algorithm, Ouroboros. Also check out Hoskinson’s rebuttal. To summarize, Larimer believes that Ouroboros, with its requirement that a node needs at least 1% of the total stake delegated to it before being eligible as a block producer, will actually be very centralized. His critique is similar to the concern I outlined above where I describe how Ouroboros is prone to centralization.
Cardano is a fundamentally strong cryptocurrency, its layered blockchain design is sound, and Cardano’s security, modularity, and upgradeability will be a large plus for Cardano when it tries to attract enterprise customers.
However, Cardano has some fierce competitors - academic rigor and peer reviewed research alone may not be enough to win against years of main net stability and gigantic network effects. That said, Cardano is not going anywhere in the medium term, and it does have a strong chance at becoming a top tier smart contracts platform. As Charles Hoskinson has said before, “It’s not a sprint, it’s a marathon.”