Optimal, a decentralized, performance-enhancing memory layer for any blockchain, raised a $ 11 million lap that invited its creators from institutions such as Harvard and mine to jump from the academic world into the commercial crypto arena.

The seed round was led by 1 KX with participation from robotic companies, finality Capital, Spartan, CMT Digital, Snz, Triton Capital, Big Brain, CMS, Longhash, NGC, Animoca, GSR, Caladan, Reforge and others.

Optimum is to build what it calls the lack of memory layer of blockchains, making the way data is stored, accessing and propagating, faster, cheaper and truly decentralized, according to a press release.

The core of Optimum’s innovation is a method of decentralized coding for distributed systems, known as Random Linear Network Coding (RLNC), developed by Muriel Médard, a MIT professor speaking at Consensus Toronto 2025.

“If you think of web3 as a decentralized world computer, people have done a fantastic job on that computer part; let’s say, the operating system,” Médard said in an interview. “But anyone who has put together a computer knows that you also need a bus that is data replanting and you need a memory that we call random access memory, as opposed to more static memory like a disk or cloud.”

Read more: Muriel Médard: Web3 has a memory problem – and we finally have a solution

Without a scalable memory layer, Blockchains are facing systemic inefficiencies, according to Médard, such as outdated gossip network, redundant propagated data, overloaded mempools, causing unpredictable delays and bloated nodes that make the retrieval expensive and complex.

Optimal’s memory infrastructure tackles ineffective data formation, redundant storage and slow access using Médard’s RLNC coding form.

Optimum is now live on a private test network and invites L1S, L2S, Validators and Node Operators to experience its decentralized memory layer in action.