Solana Lost 90% of Its Speed Going Quantum-Resistant. That's the Real Bill
A Solana test network ran about 90% slower once quantum-resistant signatures went in, and IOSCO is warning that the migration itself, not the quantum computer, is the near-term risk. Here's who pays, who gets caught with old keys, and the one date that matters.
What you need to know: a Solana test network ran roughly 90% slower after quantum-resistant signatures replaced the usual set. That's the first hard number attached to a problem the industry has spent two years treating as a conference panel.
The number that matters today is 90. Not because it's final. Because it's the first honest measurement anyone's published.
The Order Things Happened In
Start with the warning. IOSCO, the umbrella group for securities regulators across more than 130 jurisdictions, put out a paper on post-quantum readiness. The finding wasn't the scary version. Nobody at a regulator's desk thinks a quantum computer cracks Bitcoin next Tuesday. The finding was duller and far more useful: the migration itself is a risk. Upgrades can drag chains down, and gaps open up while the work sits half-finished.
Put it on a timeline. NIST finalized its first post-quantum signature standards back in August 2024. That gave every chain a target to build against. Roughly eighteen months later, most of them are still in the design-doc stage, and the ones running pilots are discovering the arithmetic.
Then came the measurement out of the Solana test environment. Same chain, same block time, quantum-resistant signatures swapped in. Throughput fell to about a tenth of baseline.
That's not drama. It's payload size. An Ed25519 signature is 64 bytes. Lattice-based schemes like ML-DSA land in the low kilobytes. Thousands of times bigger. Every transaction carries more data, blocks fill faster, and validators burn more CPU and bandwidth for the exact same 400 millisecond slot.
And the third piece, the one nobody wants to own. Wallets that keep using older keys stay exposed. You can upgrade every validator on the network and still have a chain full of accounts sitting on keys generated in 2017.
Who Actually Pays
Validators feel it first. Bigger signatures mean more storage, slower propagation, more expensive hardware. Solana's entire pitch is cheap and fast. A 90% haircut on a testnet lands it roughly in the neighborhood of the chains it spends marketing budget dunking on.
But testnets lie in both directions. They're unoptimized by design. Signature aggregation, batching, and smarter account abstraction all claw back real throughput, and none of those were in this run. So treat 90% as a direction, not a verdict.
The direction is the problem.
Here's hot take one. Slow is fine. Anyone who'd rather keep a fast chain with breakable signatures than a slower one that survives a cryptographically relevant machine is trading a permanent risk for a temporary convenience. The chain that ships slower and safer wins the decade. The one that ships fast and hopes nobody builds the hardware is renting time.
Hot take two, and this one's less comfortable. The quantum computer isn't the near-term threat. The migration is. Roughly 1.7 million bitcoin sit in old pay-to-public-key outputs, keys exposed the day a machine gets big enough. Same story for early Ethereum accounts, exchange cold wallets nobody's rotated since 2017, and every "lost" wallet that's only lost until it isn't.
So who loses? Custodians holding legacy keys. Wallet providers dragging out support for new signature schemes. Anyone whose security model amounts to we'll get to it.
Who wins? The boring companies. Signature aggregation researchers. Hardware vendors selling bigger boxes. Exchanges that publish a rotation plan before anyone forces them to.
And here's the part that should worry holders. The cost doesn't land as a headline. It lands as fees and as latency. If per-transaction cost drifts up because each transfer carries a heavier signature, that hits high-frequency onchain strategies first. Liquidations, arbitrage, MEV. The margins on those get thinner long before retail notices anything.
What Comes Next
Watch the testnet-to-mainnet path. Solana has the most aggressive performance culture of any major chain, so if anyone claws back that 90%, it's them. The levers are known. Signature aggregation and account abstraction. Neither is exotic. Both take quarters, not weeks.
The wallet layer matters more, and it's the piece nobody's funded properly. A network upgrade is a governance vote. A wallet migration is a hundred million individual decisions, and most people make theirs by doing nothing at all.
One thing to watch: the first major exchange or custodian to publish a hard deadline for key rotation. That's the catalyst. Not a paper. Not a testnet result. A date with legal weight behind it. When one big venue moves, the rest follow inside two quarters, because nobody wants to be the balance sheet still holding exposed keys when the auditors show up.
Quick hits on what I'd want before this stops being a headline. Published throughput numbers with aggregation switched on. A real cost estimate for validators running the new signature set. And at least one chain willing to name a fork date for legacy key holders, even a controversial one. The first mover here eats some pain, then owns the narrative.
So here's the question worth sitting with. If the fix costs 90% of your throughput today, what does it cost you the morning after a serious quantum announcement hits the tape? Speed gets negotiated. Compromised keys don't.
Answer that and the roadmap writes itself. The chains that start now pay a bar tab. The ones that wait send the bill to every holder on the network.
Explore More
Key Terms Explained
A way to make crypto wallets behave more like normal apps by turning every account into a smart contract.
Profiting from price differences of the same asset across different markets.
An approval term meaning authentic, bold, or worthy of respect.
The first cryptocurrency, created in 2009 by the pseudonymous Satoshi Nakamoto.