Solana’s economic dilemma
I believe in Solana, I think it is possibly the first implementation of an “ideal” blockchain. Yes, there are of course issues (including the major one I’m writing this article about) but their approach, goals and what has been achieved so far give me tremendous hope and optimism about this project.
As a proof of stake system Solana depends on validators to process transactions and produce blocks. Validators produce more blocks the more stake they have, and they must vote on every block they see in the network, for which they also get rewards, proportionate to their stake-weight (their stake relative to the total activated stake on the network).
Because Solana’s focus is on high-throughput and low-cost transactional abilities with built-in smart contracts coded in Rust and loaded into their BPF loader, the hardware required to operate a validator is on the high end.
A minimum of 12 cores, preferably more, with 3GHz base clock and 128GB RAM as well as 1TB in NVMe drives is a minimum to be able to keep up with the speed of the chain on their mainnet.
Solana operates in units of time called epochs, which are groupings of blocks, at epoch boundaries the network changes state of stake accounts and recomputes the leader schedule of validators chosen to produce blocks in the upcoming epoch.
An epoch consists of 432,000 blocks, targeted to be 400ms apart. Validators are supposed to “vote” on every block they observe on the network (via the gossip protocol). Each vote is an on-chain transaction and attracts a fee, currently this is about 5000 lamports (1 lamport being equal to 0.000000001 SOL). This results in an average daily operating cost just in terms of voting fees of about 1 SOL.
Economics of a validator
So what are the economics of a validator?
Costs
Hardware & networking costs, hardware as per above spec is quite high-end, network requires 1gbps with 50+ TB per month in bandwidth (more stake requires more bandwidth). Top quality servers costs upwards of $500 and finding providers able to provide them can be difficult (we use Maxihost).
Voting fees: 1 SOL per day
This means a monthly cost (at current SOL prices) of over $5,000.
Rewards
Validators earn rewards for voting as well as for producing blocks. Vote rewards are paid by inflation and thus independent of the transactions fees, while block rewards consist of 50% of the fees of the included transactions (the balance being burned).
Rewards can be calculated as:
R = I * C * S + B * S — V * E
Definitions:
R: Rewards
I: Annualised inflation
C: Validator Commission
S: Total activated stake
B: Block rewards (current observed estimated are 0.038% of total stake)
V: Voting fees (approx 1 SOL per day)
E: Epochs per year (approx 146)
To break even a validator must hold a minimum of around 5,000 SOL in self-stake, this is their own SOL they have staked, since they can keep 100% of rewards for this, alternatively they must have around 50,000 SOL in third-party stake with 10% commission (and more third-party stake the lower their commission is).
In fiat terms this means a validator must have $800,000 worth of SOL, or third parties must stake $8,000,000 worth of SOL with them. Compared with Ethereum’s requirement of 32 ETH to run a validator which is approximately $110,000.
Solana’s economies do not account for a massive increase in the price of the native token, six months ago the price of SOL was just 10% of what it is now, and thus the costs of launching and operating a validator have skyrocketed in fiat terms. This is a major headache for the chain that is trying to improve it’s decentralization and censorship resistance.
If this isn’t addressed soon the growth of the validator community will stall, already large “whale” validators that control 5m+ in stake continue to grow with no incentive for delegators to choose smaller nodes, while the barrier to entry has become impossible to overcome for many.
Discussions around reduction of voting fees have been ongoing among the validator community, a discussion in which we have also been very involved. This article is intended to summarise and shed light on the situation as is, and we’ll post a follow up proposal for a reduction in vote fees in the coming week or two.
If you’d like to support our validator, you can stake with us, simply search your wallet for “Laine” or get our vote account and other relevant info from the page on our website.
