Thread by @CryptoSeq

Part Two:

1/ Why does Avalanche $AVAX stand above the rest?

✅Security
✅Performance
✅Decentralisation
✅Customisability
✅Tokenomics

Next up is Performance and how Avalanche compares to other platforms. Link to the 1st below and then separate threads for the others later

https://twitter.com/CryptoSeq/status/1380862159643672578

2/ Classical consensus protocols are based on all-to-all voting and typically have a designated leader who initiates the decision process and a series of rounds of all-to-all communication to ensure that all correct nodes reach the same decision.

3/ They typically require quadratic communication overhead with all-to-all communication of O(n²). This means for each round if there is:

A network of 10 nodes = 100 messages
A network of 1,000 nodes = 1,000,000 messages
A network of 100,000 nodes = 10,000,000,000 messages

4/ So an easy way that many blockchains cut down on that overhead and achieve better performance is simply to sacrifice decentralisation and only run a small number of nodes.

https://twitter.com/CryptoSeq/status/1364508841895944192

5/ Binance Smart Chain is just a clone of Ethereum with only 20 validators and the use of a centralised Proof of Authority consensus and where the majority of nodes are controlled by Binance.

Achieving better performance but completely centralised...

6/ Avalanche Consensus is the biggest breakthrough since Nakamoto. It uses repeated random sub-sampling of the entire network to quickly achieve consensus with minimal overhead per node for incredible performance. See this article for more details
medium.com/avalanche-hub/…

7/ The number of messages each node has to handle per decision is O(k) and does not grow as the network scales up to millions of validators all participating in consensus to achieve truly global scale decentralisation for permissionless blockchain to take blockchain mass adoption

8/ Another easy way to achieve higher performance is to increase the hardware requirements for running a node, the more hardware resources, the higher the performance, but yet again at the expense of decentralisation.

9/ Solana recommends validators run hardware with 24 Cores, 128 GB Memory, 4 TB NVMe disks, Nvidia 2080 ti graphics cards (4000 cores) and very high bandwidth connections. This limits those capable of running a node though due to the expense, similar to Bitcoin ASIC mining rigs.

10/ Solana's claimed tps numbers are also vastly inflated due to consensus voting messages being included on-chain. The more validators there are the more votes there will be, and fewer available for actual transactions.

11/ If Avalanche were to include consensus messages in its claimed tps then it would be capable of 1.3 Million tps for just for the primary network.

Solana's tps is achieved through being able to run parallel execution using the 4000 cores of the GPU.

12/ The problem with this though is that it works when you are processing the same computation simultaneously. Which is great if you have a blockchain with a few smart contracts as those smart contracts will be frequently run and can be used by the GPU to process them in parallel

13/ But as the ecosystem becomes diversified with lots of different smart contracts / computation then the GPU's can't be used and instead has to use the CPU's instead for which there are far far fewer cores. The more diversified the ecosystem is, performance degrades drastically

14/ Avalanche is a platform of platforms where blockchains can be built on top. In the scenario above application specific blockchains, which are tailor made to meet the developer's requirements can be used.

15/You can have a subnet which requires high validator requirements and as it's specific to that Application, it can benefit from GPU processing without being slowed down. Whereas Solana is a single blockchain trying to cater for every use case and developers having to compromise

16/ With Avalanche there can effectively be many Solana type blockchains build on top of it, each benefiting from the full high tps that solana claims, whilst being specifically designed for their DAPP. Rather than 1 Solana which everyone uses, you effectively can have 1000's.

17/ Rather than forcing high validator requirements for everyone though, the primary network on Avalanche requires just 2 cores, 4 GB Memory and 32 GB Hard Drive space and yet can scale to 4500 tps per subnet. If you need more performance, then subnets can provide that.

18/ Other platforms also accumulate their tps over the number of shards, I.e. each shard can do 1000 tps and can scale to 100 shards, therefore 100,000 tps..

Avalanche can do 4500 tps on just 2 cores, 4 GB memory, whilst also able to scale to unlimited number of subnets.

19/ Speed isn't everything though, what's equally important is latency. It's pointless to say you can process more transactions that VISA and take 60 seconds for the transaction to be final. Can you imagine buying an item at a store and waiting 60 seconds for the payment to clear

20/ Do you like waiting?

Time it takes for a transaction to be considered final:

Optimistic rollups: 1 week
Bitcoin: 60 mins
Ethereum 2.0: 6 mins
Cardano: 5 mins
Polkadot: 60 secs (60 mins to
external chains)
Elrond: 51 secs
Solana: 13 secs

Avalanche: less than 1 second

21/ Everybody nowadays expects things to happen instantly, providing a superior user experience. With Avalanche that's possible. It's not just for payments but effects the user experience for all applications. Nobody likes waiting.

22/ Layer 2 solutions can provide additional performance but not without their disadvantages. What's important is to have a scalable layer 1 solution to start with such as Avalanche from which layer 2 solutions are also compatible.

23/ Each Layer 2 solution fragments the network, (there are even multiple separate chains within each different layer 2 solution). Rather than start with a blockchain that does 14 tps (and even ETH 2.0 with 64 shards each are only capable of eth 1.0 speeds) you instead use a ..

24/ scalable Layer 1 solution which offers higher security guarantees, more decentralised and high performance of thousands of tps to begin with, then layer 2 solutions building on top of that are naturally going to offer far more tps with less fragmentation.

25/ Avalanche is a high performance blockchain, with 4500 tps per subnet (scale to far higher tps per subnet for those that require higher resources) supports an unlimited number of subnets, sub second finality and all without sacrificing decentralisation

https://twitter.com/CryptoSeq/status/1364508841895944192

26/ Why compromise on one single blockchain when you can have a platform with thousands of them? each dedicated to your specific requirements if you wish. Whilst also able enable regulatory compliant subnets for enterprises to build on, potentially supporting trillions in assets