What is a SIMD?
A SIMD, or Solana Improvement Document, is a publicly accessible proposal for modifying the Solana protocol. It lives in the solana-foundation/solana-improvement-documents GitHub repository. Any major adjustment to Solana - whether it's tweaking the fee structure or revamping consensus, begins its life as a SIMD, which can be read, discussed, and submitted by anyone.
Think of it like a local zoning proposal. You can submit a new design to change a neighborhood, but it gets made public for the community, debated during comment periods, vetted by engineers, and voted on before the construction crews roll in. A SIMD is the design proposal; the discussions and vote that precede it are the point.
How does a SIMD get adopted?
It starts as a pull request, with an author submitting a spec and the debate plays out in the open. Core engineers, validators, and any other GitHub user can offer critiques. A handful of proposals fail at this stage, others are completely overhauled, and the entire thread remains open for review. A design that survives gets implemented inside feature flags, which allow the code to ship with a validator software release but remain disabled by default until specifically turned on. Validators choose when to upgrade to a particular release, and the feature becomes enabled once a critical amount of stake upgrades.
More significant economic changes must first undergo validator governance votes. Proposals that concern issuance, fees, or reward distribution are voted on-chain, with validators voting on behalf of their stake. Votes span multiple epochs, which typically take days, to allow sufficient time for deliberation. Tooling released in mid-2026 improved this process further by allowing delegated users to override their validator's choice and cast their own vote in governance. So now, delegating to a validator no longer means giving up your vote on questions around the protocol.
Important SIMDs
Let's take SIMD-0326, better known as Alpenglow. That's a proposed update to Solana's consensus protocol, which passed in September 2025 and was approved by more than 98% of voting stake. SIMD-0286 increased the block capacity limit to 100 million compute units (the "unit of measurement" for a transaction), meaning there would be more space per 400ms block. SIMD-0123 and SIMD-0185 shifted revenue flow so validators automatically distribute priority fees to their delegators, directly affecting how much you'd earn staking your 100 SOL (~$8,000 at ~$80/SOL).
Perhaps the most valuable example is the one that failed. SIMD-0228 was a proposal to replace the current Solana fixed inflation schedule with market-based emissions. It was rejected by validators in March 2025. This failure is just as important as any of its successes because a voting mechanism that can reject a well-publicized proposal is actually doing its job, rather than rubber-stamping everything.
Why SIMDs matter
If you've used Ethereum, the concept should feel familiar- these are Solana's answer to Ethereum Improvement Proposals, the way Bitcoin Improvement Proposals are to Bitcoin. The distinction comes in how each network ultimately resolves proposals. While Ethereum resolves through general community agreement among client teams and the community, without any formal on-chain voting, Solana puts its most important economic proposals in stake-weighted on-chain votes. Whether that makes the process more valid or just more plutocratic is certainly debatable and one of the more honest criticisms against the system is stake-weighted votes give disproportionately high sway to validators and staking pools, and thus just because the proposals are written in public does not mean they are decided in public.
For your own bottom line, the decisions are being made now. The block rewards you are receiving when you stake is because a few SIMDs have passed. There's currently a proposal for a change in the inflation rate (SIMD-0550), which would drastically alter how much you earn in your staking account. When you come to realize your yield is 2.1% instead of 1.6% next year, there's a good chance that's the reason. Tracking things down doesn't cost much. The repo is public, Helius publishes plain language write ups of larger proposed changes, and Solana Compass keeps a running tally of votes as they happen. Taking a quick glance at a SIMD when it's up for vote is probably as close as we can get to reading the bill before it became a law.
Here's a concrete example: you delegate 500 SOL to a validator (~$40,000 at $80/SOL) and SIMD-0550 is up for a vote to speed up disinflation. By reading the spec, you know exactly what yield is affected by the change; by checking with the validator, you know how the stake will be cast. Choosing to not participate allows decisions on your behalf to be made without your input.
Do SOL holders have any say in SIMDs?
SOL holders do have some say, but there are some caveats. Anyone can leave comments on the GitHub proposal, and governance votes are voted on-chain by validators weighted by delegators- with mid-2026 features, stakers will be able to override the validator's vote directly so only SOL holders who do not delegate have no say at all.
