Comprehensive dismantling of AVAX: from consensus to subnet dynamics and project sharing

Surprisingly, while the core operating mechanisms are very simple, these protocols lead to highly idealized system outcomes that make them suitable for large-scale deployment.

 ——Avalanche Platform White Paper

The battle of the public chain is still going on. The recent creation of ATH, Luna, and the betting protocol of the founder Do Kwon have all made Terra shine. Cosmos is still very strong in a bad market environment, and at the same time, it has also gained a lot of attention by relying on a large number of airdrop expectations.

As a T1-level public chain Avalanche, it seems to have lost its voice recently.

Although there is a lot of discussion about the subnet, it seems that the market does not have a clear understanding of the potential of the subnet or the vision of Avalanche itself.

The $290 million Multiverse incentive plan released by Avalanche recently has some parts that have been ignored by the “market”. As a leek that has been concerned about Avalanche for a long time, we believe that these neglected details just prove Avalanche’s firm vision and development potential.

The content of this issue will start from the principle and share with you what is Avalanche, what is the subnet, some recent market actions of Avalanche and several projects worthy of attention, which only represent personal opinions, not investment suggestions. Welcome everyone to join the Avalanche community at the end of the article. Communicate with us.

The content of this issue will not involve the token model, lock-up situation, etc. This part of the content comes from Galaxy Digital’s article: Galaxy Digital Research’s analysis of AVALANCHE has been very clear. It is recommended to take it together with this content for better results.


What is Avalanche

Snowball+DAG, Avalanche is an open platform

Avalanche defines itself as: an open platform suitable for deploying Dapps and enterprise-level blockchains. Also because Avalanche is suitable for many blockchain deployments, when you compare L1 public chains, in the long run, it is more appropriate to compare him with multi-chain parallel projects such as Polkadot and Cosmos.

Of course, in terms of its own public chain performance, Avalanche’s transaction speed and degree of decentralization are relatively good. Even in the case of node doubling, the transaction speed and security on the chain will not be affected. Compared with COSMOS, in addition to Avalanche’s efforts to bring traditional finance to the chain, which will be described later, since the creation of each subnet requires 2,000 AVAX to be pledged, the value capture of Avalanche’s native token is also higher.


The core difference between Avalanche and other public chains is the consensus protocol, and we will also expand from this.

Avalanche’s Consensus Evolution

Slush->Snowflake->Snowball->Avalanche, from simple repeated sampling to complete consensus.

Consensus refers to the process by which a series of independent voters (ie: validators) agree on a decision.

There are two main types of consensus protocols in the past:

1. Classical Consensus Protocols

2. Nakamoto Consensus

However, the two have made compromises in terms of scale, transaction speed, etc., and there is the impossible triangle of blockchain that we often talk about later. That is: decentralization, high performance and high security cannot be satisfied at the same time.


The consensus mechanism of Avalanche claims to be able to break the limitations of the impossible triangle of the blockchain. The following is a comparison of different consensus mechanisms officially given.


The consensus formation of Avalanche is mainly divided into four stages, Slush, Snowflake, Snowball and finally the Avalanche consensus protocol formed by the combination of Snowball+DAG.


Before sharing the consensus evolution process in detail, friends who know Avalanche may have heard a word called: metastability.

Metastability is when the flip-flop cannot reach an identifiable state within a specified period of time. This is also related to Avalanche’s continuous subsampling, and we start with Slush.

Stage 1 Slush: Introduce metastable state and perform simple repeated sampling.

This is the foundation on which Avalanche has evolved. Slush is inspired by the Gossip Protocol (or: Epidemic Protocol), the gossip protocol that Bitcoin uses to broadcast transaction and block information.

A visual analogy for the gossip protocol is: imagine the process of gossip spreading or the process of eating melons. Often, a person first tells you “what melon?”, and then you “spread this melon” to other people. People, in the end everyone “knows this melon”.


The picture comes from Zhihu user @juniway

Slush has optimized the above “process of eating melons”, that is, by constantly verifying the authenticity of the melon from the people around him, and finally confirming whether he believes it or not. This process is: repeated subsampling.

At its core, suppose you have three states:


1. Uncertain state: that is, you have not been spread, and you do not know what melon is;

2. Believe in this melon: that is, you believe that this melon is true ;

3. Do not believe this melon: that is, you believe this melon is fake.

First of all, as the unknowing you, you will complete the consensus through the following process.

a)  start sampling with the expectation of believing or not believing;

b)  The sampled nodes return whether they believe or not;

c)  According to the principle that the minority obeys the majority, if the majority result is believed (such as believe: do not believe = 3:2), the sampling node chooses to believe, if the majority result is not believed (such as believe: do not believe = 1:4), sample Node chooses not to believe.


The node graph is provided by Zhihu user @JoeQuant-Jackal

For safety, multiple sampling will be performed. If the results of several consecutive samplings are consistent, the node will finally change its state. The dynamic sampling process is as follows:


The picture is provided by Zhihu user @JoeQuant-Jackal

How many nodes are selected and how many consecutive samples are required to be consistent, that is, the parameters k and α mentioned in the white paper.


But this sampling process actually encounters a problem.

That is, if a malicious node adjusts itself to the opposite state, so that the sampled node cannot complete the final confirmation in the correct state, the network security will be insufficient.

Based on the above problems, there is the concept of SnowFlake, and the concept of a counter is added to Slush.

Stage 2 Snow Flake: Add a counter to Slush to record the number of times the nodes have reached consensus in history.

A feature of Slush is: memoryless, that is, the node will only retain the final state, but will not record the sampling history. SnowFlake, on the other hand, will allow the node to save the “consistent number of previous consecutive samples”. The specific implementation is as follows:

– Increment a counter for each node;

– If the sampling result is different from the previous round, the counter is reset to 0;

– If the sampling result is the same as the previous round, the counter +1;


Finally, when the “number of consecutive sampling consistency α” is greater than “a certain value β specified by the system”, the status confirmation is completed.

The advantage of this is that even if there are wrong samples in the middle, if the accumulated α of the previous sampling results of the node is greater than β, the state switching can still be completed.

However, this confirmation process still encounters a problem, that is, if the malicious nodes occur frequently, the counter will repeatedly return to 0, so that some nodes cannot reach a consensus, so the sampling has been repeated.

Based on the above problems, there is the concept of Snowball, which improves the counter in Snowflake to the concept of confidence.

Stage 3 Snowball: Add confidence to Snow Flake to measure the historical verification quality of nodes.

The core problem of Snowflake is that nodes do evil and make the counter reset to zero repeatedly, eventually causing the network to fail to reach a consensus.

After the improvement to “confidence counter”, the node will not change the state or reset the counter to zero because the single sampling is different from the previous one, but reduce its own confidence, and the final color change will be determined according to the value of the confidence.


This is where the Snowball consensus comes from and is one of the cores of Avlanche.

Interested friends can also experience the formation process of the Snowball consensus, and drag the mouse in the matrix diagram to simulate the evil node.


And Avalanche has also made a layer of upgrades on this basis.

Phase 4 Avalanche: Add the concept of DAG to Snowball to increase the efficiency and security of transactions.

In order to make the network more efficient and secure, Avalanche also added the concept of DAG to Snowball.

DAG: Directed Acyclic Graph, directed acyclic graph.

The data structure of the blockchain is a linked list (not extended yet), which is a linear structure, while the DAG is a graph structure, so that transactions can be performed in parallel to speed up transactions.


The picture comes from knowing user Peter Wang Guangzhong

Another feature is that because each transaction has a directional arrow (directed concept), the parent-child relationship between transactions will be intertwined with each other, so if you want to tamper with a transaction, it will increase the complexity of tampering. The cost of doing evil will increase.


So Snowball + DAG is what we call the Avalanche consensus protocol. It should be noted here that the three chains of Avalanche are not all Avalanche consensus.


Because the data of the P chain and the C chain is still in a chain structure, the Avalanche consensus can only be used in the transaction scenario of the X chain. The P chain and the C chain use a linear consensus Snowman customized based on Avlanche.


Based on the consensus of Avalanche and Snowman, Avalanche has officially conducted actual measurements. The data shows that even when the number of nodes increases to 2000, the throughput will not change much. Even if there are more nodes in the future, the avalanche consensus is still done in this way of “repeated subsampling”, so theoretically the network will still maintain a faster speed.


The above is the core introduction of Avalanche in terms of consensus. Of course, there are some details in the white paper, such as how to quickly deal with some transaction conflict scenarios, the impact of node changes on latency, how Avalanche views sharding and other issues. .

We have made some simple annotations in the white paper. Interested friends can reply to “AVAX” in the background of the WeChat public account to obtain our simple annotated version of the white paper and enter the community.

Relevant concept distinction of Avalanche

The relationship between X/P/C chain, Primary Network, validator, subnet and blockchain.

The more common diagrams on the market are as follows. The diagram introduces the structure and characteristics of the Primary Network in detail.


simply speaking:

X chain: mainly used to create and trade assets;

P-chain: carries the metadata of the Avalanche network and is used to coordinate validating nodes and create subnets;

C chain: This is an EVM-compatible chain used to create EVM-related contracts, etc.

It should be pointed out here that only the X chain is the Avalanche consensus, so the X chain belongs to AVM (Avalanche Virtual Machines). At present, the most common scenarios used by users are: interacting with the Avalanche wallet and the exchange wallet, but in fact this does not mean The potential and wide range of use cases of X-Chain.

Avalanche has a vision to bring more traditional financial assets to the chain , which requires the definition of assets. For example, this asset can only be traded by people in a certain country, or can only be traded in a certain period of time. Or trade in other customized scenarios, etc.

The definition of X-chain is: a decentralized platform for creating and trading digital assets. Such a feature would have the opportunity to fulfill Avalanche’s vision and is one of the features of Avalanche that is widely ignored by the market.


As there are more and more discussions on subnets, related concepts such as the validators involved in the subnet and the relationship with the X/P/C chain need a more complete picture to sort out.

We summarize the relationship between Primary Network, X/P/C chain, subnet, and verifier as follows. There are several pre-concepts that need to be clarified here:

1. A subnet is a network composed of a series of validators to reach consensus on the blockchain;

2. Each blockchain can only be verified by one subnet;

3. A validator on each subnet can verify multiple subnets;

4. Each subnet is a member of the Primary Network and needs to pledge 2,000 AVAX.


The figure lists three Subnets1/2/3, and they are verified by a set of multiple verifiers A/B/C respectively.

Go deep into the subnet and explore the rules and potential of the subnet.

As can be seen from the above figure, each Subnet is a member of the Primary Network, and the P chain in the Primary Network serves all the subnets, which is why it is said that the custom subnets are customizing their own areas. At the same time as the blockchain, it still enjoys the guarantee of the overall network of Avalanche.

At the same time, as a subnet, you can see that Subet3 can verify Subnet2 at the same time, but it does not need to verify Subnet1. That is to say, each customized subnet can only pay attention to/verify the network data that it is interested in, and does not need to verify the network data that it is not interested in, thereby increasing the burden on itself.

The above are the benefits brought by the structural characteristics of the subnet.

At the same time, the core potential of subnets is that you can customize the rules of the network to make this chain more suitable for your business. For example, some of the assets mentioned above can only be traded by people in certain countries. Similarly, you can also allow nodes on your network to be accessed only by devices in certain countries, or use other restrictions to create your own blockchain.


For example, for the blockchain of the game, you hope that the hardware configuration of the verification node is relatively high, and you can also put forward requirements for the relevant verifier.

The two most popular game Subnets are Swimmer Network led by Crabada and DFK Chain led by Defi Kingdom. While improving network speed and providing incentives, both of them also use their respective main tokens as new public tokens. The gas fee of the chain is combined, and the usage scenario (consumption) of the native token is increased.


While customizing the blockchain, you can also customize the virtual machine, such as Subnet3 in the figure above. At present, Avalanche’s C chain is mainly compatible with EVM. In theory, developers can customize various types of VMs (Virtual Machines) through Avalanche, and even use Go language.


At present, there are not many projects with subnet plans on Avalanche. In addition to Crabada and Defi Kingdom mentioned above, there are Ascenders, Shrapnel, Cryptoseal, etc. Most of the projects are related to games and are still in the development stage.

Although Avalanche has previously worked with companies such as Deloitte on custom blockchains, for Avalanche, Subnet’s potential has yet to be truly unlocked. The current number of subnets on Avalanche, interested friends can continue to follow up.


Avalanche’s new progress and project introduction

Hackathon competition and Multiverse incentive program.

Avalanche is currently working on several major events are the Summit in Barcelona, ​​Asia’s hackathon program, and the latest $290 million Multiverse incentive program.


The above actions are all creating fresh blood for the Avalanche ecosystem. There is no more introduction about the hackathon and Multiverse here, and interested friends can check it out by themselves.

The main thing I want to mention here is something that has been widely ignored: in Multiverse’s plan, there is a paragraph that says that Avalanche will provide institutions with native KYC functionality on the chain.


Such a function happens to be another important move for Avalanche to bring traditional finance to the chain.

Avalanche’s Mainstream Projects Related to Subnets

We share in Gamefi/Defi/NFT/DAO.


1. As the most user-friendly chain game on Avalanche, Crabada has launched the test of its sub-network SwimmerNetwork;

2. Defi Kingdom is the first game sub-network project in the Multiverse plan;

3. CryptoSeal is committed to building Loot on Avalanche, and the subnet has also been tested online;

4. Ascenders has just started the demo beta test of some of the game’s gameplay recently, and its delivery quality is relatively high;

5. Wildlife is one of the largest mobile game developers in the world, and its development progress of its sub-network is still worthy of attention.


Defi (there is currently no officially announced sub-network project, welcome to add)

As the native leading Defi of Avalanche, Trader Joe ‘s user volume/TVL/profit has always been in the forefront. It is worth mentioning that after Trader Joe changed its economic model, the value capture of its JOE is also improving. At the same time, it is suspected that the official will Further release of NFT mall. If Defi is going to have a subnet, Trader Joe’s is more likely.


NFT (there is currently no official announcement of the subnet project, welcome to add)

1. Kalao is Avalanche’s native NFT trading platform. Almost all Avalanche NFTs will be listed on Kalao. At the same time, Kalao also provides VR-type displays. Kalao may be one of the initiators of the NFT sub-network project.

3 HopperGames is the most recent NFT in Avalanche in terms of transaction volume. The team is from PartyAnimals. For Hopper’s NFT, the team has set up a very rich gameplay. In our opinion, the project party may be able to open up some NFT gameplay and the design logic behind it. Give more projects to build their own NFT subnets.



1. Colony is a community-driven DAO owner on Avalanche (currently not enough DAOs). Although Colony may not necessarily develop subnets, it has made it clear that it will participate in the staking plan of the subnets.

2. AVentures is a relatively well-known investment DAO on Avalanche. Most of the community members are Avalanche OG. Although they mainly focus on investment itself, their status is also at the forefront of the Avalanche DAO project.


None of the above is any investment advice, only an overview of projects that may be linked to the subnet. Welcome everyone to reply “AVAX” in the background of WeChat public account to enter the community and communicate with us.


After we understand the technical composition of AVAX and the possible development of the current subnet, we will have a clearer understanding of the whole of Avalanche. Also, when flipping through the white paper, we were very impressed by this sentence:


Literally: Surprisingly, while the core operating mechanisms are very simple, these protocols lead to highly idealized system outcomes that make them suitable for large-scale deployment.

Translated in web3 language: x*y=k (the core principle of Uniswap, constant product)

In the language of web2, it is translated as: Avenue to Jane

There are new projects coming out every year, but in fact, the infrastructure is not updated and iterated as fast as everyone thinks, which may surprise you, but there are not so many fundamentally different technologies–Ted Yin|Avalanche joint founder.

Technological innovations are not easy to come by, and once they are created, their impact is beyond imagination.

What the market really needs is innovation, not various imitations. And when we research and explore innovation, the value this process brings to us will be beyond our expectations.

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2022-03-26 09:51
Next 2022-03-26 09:53

Related articles