Feeling particularly weird and impulsive today.
Not sure why, people always call us QUIRKY, but lately it’s been worse…
Today’s review is Quarkchain...stfu and read along. You'll like this one.
Let’s dig in….
Quarkchain (ticker: QKC)
Wtf is this shit?
Quarkchain is a blockchain project aiming to handle over 1 million *said in Dr. Evil voice* transactions per second (tps) while reliably being able to run decentralized applications. To achieve this, they take advantage of various sharding techniques.
Sharding is a method of splitting a large quantity of data across many containers in order to improve the speed by which that data can be accessed.. in traditional, boring, old, centralized, insecure, STUPID, OUTDATED, F***N DATABAS - woah, sorry getting ahead of myself….deep breaths.
Let’s try that again. Sharding splits up a large data set into smaller subsets. Here’s an example (n00b). A company has a database of usernames that contains 100,000,000 usernames that start with letters a-z. To query a username that starts with the letter “y” can take both time and computing power...and at a bigger picture, leads to an inefficient system for quickly finding/validating/parsing through different users.
Instead, you can partition the data up into smaller databases. One database contains users with usernames that begin with letters A-C, then D-F, G-I, and so on. With some magical dev skills, the system would know which dataset to jump to depending on the username that needs to be found. This makes for a much more scalable, fast, and efficient system. This same logic can be applied to blockchains and how they process transactional data.
Now...Quarkchain’s whitepaper clearly expresses the three main issues with blockchains currently: security issues, decentralization issues, and scalability issues. Often times these issues overlap...so it becomes a balancing effort. If you want to tighten security by creating more work required, or forcing all nodes to validate every transaction (like with Bitcoin), then you will do this at the expense of scalability. If you want to improve scalability by introducing super-nodes or delegated nodes and such, then you can, but often at the expense of shittier security. You get the point.
So...keeping all of this in mind, Quarkchain has introduced a possible solution for increasing transaction speed without hurting decentralization and security too much.
They have a two-layer architecture: the root chain and the many blockchain shards.
The sharding blockchain layer is a network of smaller blockchains (shards). Each shard processes some sub-set of all transactions on the network (this is kind of obvious if you understanding sharding). This layer is also known to be elastic, meaning more shards can be made and as a result, the tps should rise as well….theoretically. They say it is easy to add shards, but don’t go into much detail.
The root chain then confirms the blocks from the different shards. To reiterate, the root chain does not process transactions like the sharding blockchains, rather it confirms the blocks from them. It is these two layers which allow for high tps (scalability), while also maintaining high security.
^^ YO BIG PICTURE SHIT…...With current single blockchain systems, the blockchain serves two purposes: it processes/records transactions to a ledger AND it then confirms the processed transactions by mining the block via PoW (at least with Bitcoin).
Anyways, splitting up these two elements into two layers (sharding layer and the root chain layer) is the real reasoning behind their architecture. *mind blown*
Moving on...decentralization. They want to discourage the creation of mining pools by distributing hash power across shards, so each shard should be receiving a similar amount of work done to it. Additionally, they encourage cheaper nodes (smaller miners) to group together and create super-nodes, which yields more rewards….we assume there must be a limit to the number of nodes that create a cluster node....otherwise it’s like a mining pool. lol
ALSO, they claim to avoid the 51% double spend attack issue through this two-layer architecture. Without getting too gritty, you would need to revert transactions on the sharding layer as well as the root chain layer which does have substantial difficulty but is not economically efficient. The root chain itself has 50% more hashing power than the sharding layers themselves...so...lots of hashing power required, thus, not economically feasible at all. What they fail to mention in their whitepaper is that unfortunately, there are many 51% attackers who attack blockchains only to hurt the reputation and not for financial incentive.
Consensus wise, they have a broad process of consensus where both layers communicate. Not sure if the algo is confirmed for the root chain, but they will be using PoW...and the shards will use a root-chain-first PoW. Not much info is given other than how it helps prevent 51% attacks and protects transactions.
A BIT MOAR INFO they also intend to support off-chain transactions and dApps as mentioned.
They are also another project launched as an ERC-20 token with plans to do a swap whenever the mainnet launches (they said Q4 2018..soo...soon?). We're not big fans of this strategy.
Welp, hope you get the picture...they have two layers...shards that process transactions, and a root chain that confirms them.
Who tf is behind this shit?
They have experienced ex-Google boys and people from other major companies. The founder is Qi Zhou, who got his PhD from Georgia Insitute of Technology (smarty pants). He is a former Google employee and has 15+ years of development experience.
All in all, mid-tier shitcoin...they have lots of enthusiasm, but also promise a lot. Until sharding is implemented on a system that actually has users, this is all theory. They claim to be able to handle 1 million tps, but we believe this was in a closed environment...testing environments are way different then real ones.
Hope you learned some shit and remember to give us a follow on Twitter!
- Mike and Aaron