A high throughput blockchain whose nodes perform real distributed computing tasks (Proof of Useful Work), instead of solving fictitious math problems (Proof of Work) to achieve consensus.
The blockchain world is seeing an influx of new projects in two important categories: faster blockchains and distributed computing. Ankr wants to build a network that is both. The core idea is replacing the Proof of Work consensus mechanism, which uses computation power to solve fictitious math problems, with Proof of Useful Work, which uses computation power to process real distributed computing tasks.
The end result is a blockchain that is as secure as one using PoW, but is useful instead of wasteful, as energy and spare compute goes towards solving real computational tasks. This places Ankr in direct competition with projects like Golem, SONM & others, as they too use distributed computer networks to act as one giant supercomputer.
Nodes in the Ankr network will have to be equipped with CPUs capable of executing instructions in a Trusted Execution Environment. This includes any Intel CPU since 2015, which make up 65% of the market, according to Ankr. They also plan to expand towards ARM mobile CPUs equipped with TrustZone, and Nvidia’s GPU’s based TLK. It is therefore reasonable to expect mobile and GPU based participants to the network, further down the line.
Trusted Executed Environment basically means that code is executed in a separate memory area that not even the host operating system can access. This means nodes performing computations for third parties have no means of reading that data, and therefore the respective third parties gain an additional level of trust in outsourcing data processing.
Ankr will feature a native Oracle function, it too benefiting from the trusted execution environment. It will fetch external data and feed it into the respective smart contract, without the host being able to read any of it.
dApps can be developed and run on side-chains, with the benefit of having direct access to a lot of on-chain computing power. Anker will likely start as an Ethereum fork, so Solidity could be used, but information on this is still unclear.
The Ankr token will be used to pay for running computations on the network, and earned by participants for lending their spare resources. Tokens are earned continuously by every participant for the work they provide, with an added bonus if they’re also selected to mine the next block in the chain.
What I like:
- A fitting combination of high performance blockchain & distributed computing
- Highly competent, albeit young, team, complemented by senior advisors
- Vetted by a lot of notable investment firms, such as NEO Global Capital and Pantera
- Proof of Useful Work, native Oracles, Plasma, sidechains, prototype around the corner
What I dislike:
- A lot of competition for spare compute resources – Golem, SONM, iExec, DADI, HyperNetwork, Uranus, Akash, RRnetwork, others
- Whitepaper could be more explicit (also, that stupid peep-hole for reading it through!)
- Plasma, Sharding, own storage network are mentioned as future work, with not a very detailed roadmap
A fast blockchain that doubles as a distributed computing network is a very strong proposal. The technical concepts put forward are compelling. Getting paid as you go, instead of staking something or wasting electricity in hope for a reward can be a powerful incentive for computers to join the network. The trusted execution environment can also be an incentive for customers to outsource their computations to Ankr. Rumors of a 30k ETH cap, coupled with big hype and notable investors make this ICO a very attractive one. I like the Ankr ICO a lot.