Solana’s 100M CU Upgrade: A Moral Audit of Performance at Scale
CryptoVault
When a network doubles down on raw performance, what does it sacrifice? The Solana Foundation’s announcement last July—raising the mainnet block compute unit limit from 60 million to 100 million—felt like a familiar beat in an endless drumroll. Faster, cheaper, more scalable. Yet, as I read the SIMD-0286 proposal, I couldn’t shake the feeling that this was less a technical breakthrough and more a quiet confession. Tracing the code back to the conscience, I see not just a parameter tweak, but a trade-off between capacity and care.
Context first. Solana has always been the speed demon of layer-1s: sub-second finality, negligible fees, and the promise of handling the world’s transactions without breaking a sweat. That narrative has been both its strength and its curse. After the 2021 outages and the 2022 crash, Solana rebuilt community trust through relentless iteration. SIMD-0286 was the latest—a proposal to increase the maximum compute units (CU) per block, effectively giving developers 66% more room to execute complex logic. It passed community review and went live on mainnet. On paper, this is a win for scalability. But as an economist who spent weeks auditing ICO contracts in Tokyo, I’ve learned that parameter changes often hide deeper assumptions.
Core insight. Compute units are Solana’s version of Ethereum’s gas—a measure of how much work a transaction or instruction requires. By raising the cap from 60M to 100M, Solana increases the theoretical capacity of each block by 66%. It sounds like an unqualified good: more room for DeFi aggregations, NFT mints, and on-chain gaming logic. Yet the real world isn’t a whiteboard. During my time working on the Neo-Tokyo Punks NFT project, I saw how "capacity" can be an illusion. If most transactions consume only 100,000 CU each, raising the cap does little—the bottleneck moves elsewhere (like network propagation or validator CPU). The upgrade only shines if the average transaction complexity also rises. And that’s exactly where the risk hides.
Based on my experience analyzing token distribution models, I’ve come to distrust parameters that assume uniform usage. The new cap could enable a surge in high-CU operations: complex MEV bundles, atomic multi-protocol swaps, or even on-chain AI inference. These sophisticated transactions will consume far more space, potentially clogging the block for simpler users. Open books, open ledgers, open hearts—but only if everyone gets a fair slice of the block. The first casualty of this upgrade might be the average retail trader’s transaction, which now competes for block space against heavyweight bots.
This is where the contrarian angle bites. The prevailing narrative is that more capacity equals more freedom. I argue it can also mean more centralization—of value and power. Consider validator hardware. Solana already demands high-end machines (128GB RAM, fast SSDs). A larger block means more data to verify, more memory to store, and tighter deadlines for propagation. Smaller validators, running consumer-grade gear, may find it harder to keep up. Gradually, the validator set shrinks toward well-capitalized institutions. We’ve seen this movie before in Bitcoin’s mining centralization. Building bridges where others build walls is the ethos of Web3, but this upgrade inadvertently builds a wall around the minimum hardware requirement.
And then there is MEV. Larger blocks with more complex transactions are a gift to searchers who can exploit slippage and ordering. During the 2022 bear market, I watched a DeFi protocol lose 40% of its LPs in one week because of rampant sandwich attacks. Raising the CU limit without parallel MEV mitigation is like increasing highway speed without installing guardrails. Chaos is just creativity waiting for structure—but structure requires intentional design, not just bigger numbers.
Still, I must acknowledge the pragmatic side. Solana’s core developers are among the sharpest in the industry. They know these risks. The SIMD process is transparent, and the upgrade was rolled out responsibly. My critique isn’t that this was a bad move—it’s that we, as the community, must treat it as a beginning, not an end. The audit is not the end, but the beginning. We need to track real-world metrics: average CU per transaction, validator dropout rates, and MEV profitability. Culture is the ultimate consensus mechanism—and culture demands accountability beyond code.
So where does this leave us? The upgrade solidifies Solana’s position as the high-performance L1, but it also raises the stakes for governance. I’ve seen too many projects chase capacity at the expense of fairness. Literacy in the blockchain age is power—not just understanding what the upgrade does, but what it doesn’t fix. In a sideways market, where every basis point of efficiency matters, the real test isn’t how many CUs you can pack into a block, but how many hearts you can keep aligned with the protocol’s values.
As I reflect on my years in Tokyo—learning from the failure of ChainLit, rebuilding after the crash, and now working with institutional clients—I believe the next frontier is not just scalability, but sustainable inclusion. Solana’s 100M CU limit is a step forward, but only if we ensure the ladder isn’t pulled up behind us. Open books, open ledgers, open hearts—let’s keep them all open.