Algorand v3.9 Major Update details - State Proofs, Higher throughput, Faster finality

Instead of going over usual r_CC fluff like governance and marketing/sports partnerships, let's talk about tech.

v3.9 Update:

Algorand is preparing for their v3.9.x update. This is a major update that introduces State Proofs, shorter finality, larger blocks, and AVM v7.

There was a post by an Algorand Foundation Community Manager for a Mainnet target release date of Sept 8th on Twitter (which I've seen). Obviously target dates is still just targets, so I'd treat it with a grain of salt.

State Proofs are the most exciting part of the updates but also the hardest to understand, so I'll leave that discussion for last.

The Updates

Slightly shorter deterministic finality: ~4.3 to 3.9s

• Algorand is already among the fastest-finality deterministic blockchains, so decreasing it to 3.9s is just icing on the cake.
• (Avalanche, Fantom, and NEAR are the only big decentralized ones I can think of that are faster. Others like Solana are probabilistic.)

Larger Blocks: Increases max block size from 1MB to 5MB

• This will increase max throughput (including "Layer 1" smart contracts) to ~6K TPS
• This is more for future-proofing since Algorand rarely sees above 100 TPS of real activity. (Peak activity was ~350 TPS in Dec 2021, but nowadays, 10-30 TPS is the most common.)
• There is a future plan to increase throughput even further with Block Pipelining by ~5x, which is something that Cardano is also exploring, and similar to how Solana functions natively. But it'll probably not be needed for a very long time given lack of activity/demand. Block Pipelining works by allowing the network to begin working on the next blocks (e.g. 4 blocks) concurrently before the current block is finalized. As long as those blocks don't have errors or conflicting transactions, they will all be included. Current stats show that only 1% of blocks fail to stage successfully, so block pipelining should work 95% of the time.

Algorand Virtual Machine (AVM) v7

• Nothing important for end users, so don't ask me for details since I'm not an AVM expert.
• Support for randomness oracles (but not on-chain)
• Unfunded accounts (no ALGO) can issues transactions (of other tokens) as long as fees are covered by others in the transaction group

State Proofs:

From a tech perspective, State Proofs (SP) are the most exciting part of this update. In their most basic form, these are a small certificates that prove that a sufficient (i.e. supermajority) total stake of Algorand has attested to some value (e.g. the state of the Algorand blockchain). SPs are SNARK-friendly, meaning that they can use SNARKs (e.g. what Ethereum's L2 zk-SNARKs use) to provide high compression. This allows them to be verifiable even ouside of Algorand on other blockchain's smart contracts, which important for fee-expensive blockchains like Ethereum.

State Proofs have 2 main goals:

1. Interoperability: Provide more-secure interoperability between Algorand and other blockchains, DLTs, and clients.
2. Increased Security: Provide more post-quantum resistance

Interoperability

State Proofs allow for light clients, easier node onboarding/catch-up, and cross-chain compatibility.

X-Chain Interoperability

SPs help resolve the issue of relying on untrusted off-chain oracles for cross-blockchain information and communications. Over the past years, so many bridges have been exploited due to bad oracles. Solutions such as Algorand's upcoming State Proofs, Ethereum's future Verkle Trees/thin clients, Polkadot's XCM, and Cosmos Hub's IBC are all meant to provide secure cross-chain communications to alleviate trust issues with off-chain information. (Though note that XCM and IBC only work within their respective ecosystems or with directly-compatible blockchains.)

For example, a smart contract on Ethereum could use a State Proof verifier that provides ongoing proofs of 100s (e.g. 256) of blocks of Algorand. Using compressed SNARKified SPs, smart contracts can cheaply verify the current state of Algorand, its transactions, accounts balances, etc. that a supermajority of Algorand stakes have attested to. It no longer has to rely on off-chain oracles for this information. As long as that smart contract implemented the State Proof verifier code correctly, an outsider would be able to use this in turn to trustlesslyverify transactions and states on Algorand.

There are still limitations. The biggest one is that Algorand's SPs only allow for trustless viewing the state of Algorand, which is only half of equation. Later on, Ethereum would also have to have its thin client solution from Verkle trees in order for Algorand to quickly verify the state of Ethereum on the other side. SPs also cannot be used to bridge transactions between chains or used for instant direct communications. They can be used to help bridges become more trustless, but not completely replace bridges.

New clients and Light clients

State Proofs also allow clients to quickly onboard or catch up to the current state of Algorand. Instead of having to verify the entire chain from the genesis block, which takes days to sync from scratch, they can simply verify clusters of 100s of blocks at a time using SPs of those clusters. Bridges can also use SPs to quickly check the state of Algorand without running full nodes.

Security

For security, quantum computers could break ECDSA/EdDSA protocols used to generate digital signatures and most current ZK proofs. This would allow attackers to find the private key used to generate those signatures. US Gov's NIST has been working on Post-Quantum protocols since 2016, and Algorand's FALCON protocol was one of the 4 finalists out of 82 original submissions. Attackers could look at prior transactions and their private keys. State Proofs use FALCON (in deterministic mode) and are thus post-quantum secure.

If you want more details on State Proofs, you can watch MIT's 23-minute video on it.