My TL;DR: A decentralized computing system built on top of a distributed ledger based on a DAG, which uses the novel Avalanche (by Team Rocket) consensus protocol.
All kinds of figures are flying around related to how much unused computing capacity is out there. IDC says only 15% of the world’s computing power is actually used, a study by Stanford in 2015 said more than 10 million servers are “comatosed” (powered by unused for 6 or more months), and so on.
Whatever the correct figure is, clearly there is a big opportunity to monetize unutilized compute. The increasing number of ICOs aiming to capitalize on that is a clear indication centralized computing will move from mainstream to niche. Supercomputers will move away from having an owner and a location towards being totally decentralized and globally dispersed.
Actually, two of the biggest supercomputers today are Bitcoin and Ethereum, when measuring the total computing power. Sadly, that power is used internally by participants to race and find an answer to some fictitious problem, instead of joining forces to solve real issues.
It doesn’t have to be like that, and projects such as Golem, SONM, iExec, Hypernet, Ankr, Uranus, Akash and others are in a race for your spare compute. Perlin joins the race with a few differentiators, which are very interesting to follow.
First of all, it builds upon the newly released Avalanche protocol. If you’re not familiar with it, the TL;DR of it is that a couple of months ago, a mysteriously named ‘Team Rocket’ entity released a paper that describes a consensus protocol where participants ask their close neighbors about their opinion, and update their own to fit with the majority. Every node does that, until a network wide consensus is reached, apparently very fast. Its main caveat is that it does not deal well with the double spend problem, to the detriment of double spenders, which you could even call a feature.
Perlin uses features of the Avalanche protocol on top of a DAG-based distributed ledger, to create a network that has relatively high throughput – 1.600 TPS – with Turing-incomplete smart contracts for settling computations. Perlin is not for building dApps on top of it, although it can serve as ledger for them.
The native token is called PERL and it is paid by customers for computations, earned by miners for sharing their resources, and staked by validators to verify the validity of compute that miners offer customers. Validators, who of course earn tokens for performing this task, are randomly selected from available miners.
What I like:
- First Avalanche implementation, this is a big deal
- The tech concepts are very good, and the code so far has received some great reviews
- They already released their own P2P networking stack, called Noise, claiming it’s the first proper solution for decentralized protocols
- Excellent team & advisors – they list themselves on LinkedIn
What I dislike:
- Massive competition in the distributed computing space
- No MVP yet, just some code released
- Unofficial metrics hint at 80k ETH for 30% of tokens, which would be a lot
- Very little information released, although 30k people are waiting on their announcements channel – this leads to some misinformation like “Perlin will unlock unused power in smartphones and gaming consoles” – no, it won’t, unless you badly tamper with them, which is not at all worth it, except maybe for proving a point
Whoever is going to win the race for most adopted distributed computing blockchain project, classical cloud will be the loser, long term. Sure, it will shift and become cheaper and big data centers will still exist, but pricing will not be decided by centralized entities, it will be supply and demand that will drive it. So it’s not a matter of whether these projects will succeed, but rather, which one. I’m playing it safe and betting on most of them. Perlin, with a very good team, excellent code and first implementation of Avalanche, has gotten my full attention. I like the Perlin ICO very much.