Solana's 100M CU Blocks: What Stakers Need to Know

TL;DR. On July 29, 2026, at the start of epoch 1009, Solana activated SIMD-0286 and raised its block limit from 60 million to 100 million compute units, a 66% jump in block capacity. Nothing breaks, nobody needs to migrate, and your staking rewards do not change overnight. But bigger blocks mean more fee-paying transactions per block, and fees are exactly where Solana staking yield is heading as inflation rewards shrink and SIMD-123 prepares to route block revenue to stakers. This upgrade quietly grows the pie that stakers are about to get a slice of.

Solana's blocks are now 66% bigger. The change landed on mainnet on July 29, 2026, when epoch 1009 opened with a new ceiling of 100 million compute units per block, up from 60 million. It shipped as SIMD-0286, authored by Lucas Bruder of Jito Labs and first proposed in May 2025.

Most stakers will never notice the activation itself. There are no breaking changes, no wallet updates, and no action required. So why should you care? Because block capacity is quietly becoming a staking story. Here is the plain-English version.

What exactly changed on July 29?

Solana enforces several limits on every block. SIMD-0286 raised exactly one of them: the total compute units a block can contain.

Compute units are how Solana measures the work a transaction performs. A block limit of 100M CUs simply means a validator can pack more work, and therefore more transactions, into each 400 millisecond slot.

The detail worth pausing on is the second row. The cap on how much compute any single writable account can consume in one block stays at 12 million CUs. Under the old 60M limit, one hot account, say a memecoin pool during a frenzy, could eat 20% of an entire block. Under the 100M limit, that same account can only claim 12%. All of the new capacity is parallel capacity: room for more activity on unrelated accounts alongside the hot spot, not a bigger lane for the hot spot itself.

Why did Solana raise the block limit?

Because blocks were filling up. According to Solana Foundation data, since the previous increase to 60M CUs (SIMD-0256, activated July 22, 2025), about 11.2% of blocks reached 56M CUs or more. Roughly one block in nine ran close to full over that year, and the pressure arrives in spikes, typically during volatile market moments when traders most want their transactions to land.

That pattern matters for ordinary users too. When blocks are nearly full, priority fees climb and transactions compete for space at exactly the moments blockspace matters most. Raising the ceiling to 100M CUs gives the network headroom to absorb those spikes.

The step was deliberately aggressive. An intermediate raise to 80M CUs was considered and rejected, because validator clients had improved enough over the past year to handle the full jump. This is the second increase in twelve months, continuing a pattern: track demand evidence, confirm the network can handle it, then lift the ceiling.

Was it safe to make blocks bigger?

Bigger blocks are not free. They take longer to propagate to thousands of validators and longer to replay, which can slow down nodes that fall behind. The network cleared two readiness bars before activation.

First, XDP. More than 70% of mainnet stake now runs XDP, a kernel-bypass networking mode that speeds up how validators send and receive block data. XDP is the default in Firedancer and will be on by default in the Agave v4.2 client release. Core engineers judged that coverage sufficient to propagate 100M CU blocks without endangering the 400ms slot target.

Second, staged rollout. The change went live on testnet and devnet first, behind the feature gate P1BCUMpAC7V2GRBRiJCNUgpMyWZhoqt3LKo712ePqsz, before mainnet activation at epoch 1009.

For stakers, the practical takeaway is that this was a capacity upgrade with a safety case, not a gamble with your delegated stake. Validator performance, which drives your staking rewards, was the gating factor the whole way.

Why bigger blocks matter for staking yield

Here is the part most coverage skips. Solana staking yield today is mostly inflation: newly minted SOL, currently around 6% gross network rewards and declining on schedule. The long-term plan, visible across several proposals, is to shift yield from printed tokens toward real fee revenue.

Bigger blocks feed that shift in three ways.

More fee revenue per block. Since SIMD-96 activated in early 2025, 100% of priority fees go to the validator that produces the block, with nothing burned. A 66% larger block during a demand spike means more fee-paying transactions included, which means more revenue per leader slot. Block capacity is a direct lever on how much validators can earn beyond inflation.

A bigger pie for SIMD-123 to split. Today validators keep block revenue unless they share it through off-protocol arrangements like Jito's tip distribution. SIMD-123, approved by governance in March 2025 and still pending activation as of early August 2026, will let validators distribute block revenue to their stakers in protocol, at the end of each epoch. Anza's CEO has said he expects it to ship in 2026. When it does, the fee revenue that 100M CU blocks generate becomes revenue that can flow to you.

A hedge against disinflation. SIMD-550, one of three pending tokenomics proposals, would double Solana's disinflation rate and compress staking APY faster. If inflation yield shrinks while fee yield grows, total staking returns depend increasingly on network usage, and usage needs blockspace to grow into. The 100M CU limit is infrastructure for exactly that.

None of this shows up in your rewards tomorrow. All of it shapes what SOL staking pays in 2027.

What stakers should do now

Nothing is required, but a few checks are worth your time.

Keep staking as normal. Native stake and liquid staking tokens are both unaffected by the block limit change. If you have not started, our guide to staking SOL covers both routes, and you can stake natively or get a liquid staking token through Hubra without giving up custody.

Watch your validator's fee revenue. As fee income grows relative to inflation, which validator you delegate to matters more. A validator that earns strong block revenue and plans to share it under SIMD-123 will beat an identical validator that keeps 100%, once activation happens.

Expect the next capacity debates. SIMD-0306, still pending, would raise the per-account cap from a fixed 12M CUs to 40% of the block limit, giving hot accounts more headroom. Alpenglow, Solana's new consensus protocol, is moving toward staged mainnet activation with BLS key registration already live. Blockspace and consensus are both being rebuilt at once, and each change nudges staking economics.

Do not chase headlines with your stake. A 66% capacity increase is infrastructure, not an APY event. Anyone promising immediate yield from this upgrade is selling something.

The bigger picture

Solana has now moved its block limit from 50M to 60M to 100M CUs in about two years, each step justified by measured demand and gated on proven validator performance. The direction is clear: more throughput at the same 400ms cadence, with staking yield gradually rebuilt on top of real economic activity instead of token emissions.

For SOL holders, the takeaway is simple. The network just got two-thirds more room for the activity that will eventually pay your staking rewards. You did not have to lift a finger, but understanding why it happened puts you ahead of most of the market.

This article is for education only and is not financial advice.