Thread by @CryptoSeq

1/ Solana is good and all, but it also has its disadvantages / inflated metrics which very few people seem to talk about. In addition to the answers given to @RyanSAdams and @TrustlessState I would also add the following 👇

https://twitter.com/BanklessHQ/status/1433106607810899968

2/ Whilst everyone talks about the high specifications of buying and running a machine (24 cores, 128 GB Memory, 2 TB NVMe, 1 GB Network and high end graphics card) many don't realise that Solana is the only blockchain that does consensus votes on chain)

3/ Up to 80% of tps are consumed just by validators voting on consensus. Even with the cheap cost per transaction, each validator has to spend between 1 and 2 SOL every day just voting in consensus.

4/ As the price of SOL increases so to does the cost in transaction fees of running a validator. Based on the current SOL price it would cost over $50,000 + a year just on transaction fees alone. Never mind the high power consumption, enormous bandwidth egress charges etc.

5/ Solana uses PoS and the more stake you have, the higher the probability you are selected to produce a block for a slot and earn the transaction fee / rewards.

To currently just break even you need to own at least $700,000+ worth of SOL and self delegate with 100% commission

6/ You would need far more with others delegating to you with a much lower commission.

Smaller and larger validators all pay the same transaction fee costs for voting on consensus, but as the validators with the most stake earn the rewards more frequently..

7/ you end up with smaller validators paying the larger validators which then compounds the problem creating further centralisation.

Also due to the TB's of data that is getting written to disk you could be replacing those TB+ nvme drives every 3 months due to TBW endurance.

8/ The stake distribution isn't as good as it appears either as a large part of that is controlled by the foundation which has a program where validators can apply, perform KYC and sign an agreement in return for the foundation delegating stake to their node.

9/ You also can't run an archival node and validate the chain from genesis. The early history is only stored by the foundation in google cloud and is around 20 TB since launching less than 18 months ago. Due to the bandwidth egress costs this is not made available for others.

10/ In addition to the inflated tps numbers through consensus votes, as the ecosystem becomes more diversified with different smart contracts / DAPPs then GPU's can't be used and instead has to use the far fewer cores in the CPU instead as explained in

11/ Finality time is another thing that is often confused with many thinking it's 500ms, but that is the block time. Solana achieves "Optimistic finality" normally within 5 seconds, the maximum time for all validators to have voted is 12.8 seconds for full finality.

12/ The upcoming release of EVM compatability through Neon VM whilst offering the ability to run multiple EVM instances in parallel, it also adds a lot of complexity and security risk and requires the use of a proxy server to translate between EVM and Solana

13/ As the EVM will run as a smart contract and can be upgraded, this also adds further points of centralisation / security risk if this is controlled by a few entities and can upgrade the smart contract to include malicious code.

14/ And to the remark of classifying Avalanche as just another GETH clone with no innovation, I recommend looking into Avalanche consensus as well as the Avalanche platform.
cryptoseq.medium.com/what-sets-aval…

15/ The C-Chain is just one chain within the primary subnet and there will be 1000's of subnets, support for any custom VM, regulatory compliant subnets for enterprise workloads whilst all being able to interoperate.
cryptoseq.medium.com/unlike-other-b…