Key takeaways from the fireside chat with Annabelle Huang, CEO of Altius Labs
Most users don’t think about how blockchains execute, until performance breaks. When things slow down, gas spikes, and transactions fail, the execution layer becomes impossible to ignore.
At the REDeFiNE Tomorrow 2025 event, Annabelle Huang shared why Altius Labs isn’t launching yet another L1 or L2, but is instead focused on re-architecting the execution layer, the part of the blockchain stack that many avoid touching.
Here are the key insights from the conversation.
Execution: The Hidden Bottleneck
While consensus mechanisms and data availability layers often dominate discussions, execution remains the most fragile part of blockchain systems. In practice, it’s usually the first point of failure when real-world usage intensifies.
Altius approaches this challenge by redesigning execution without changing consensus. Instead of starting from scratch, they preserve compatibility with existing virtual machines while radically improving performance under the hood.
Scaling Without Rewriting the Stack
Many projects attempt scalability by rewriting their entire architecture launching new chains, inventing new languages, and building bespoke tooling ecosystems.
Altius takes a different route: it retains the existing environments (like EVM, SVM, and MoveVM), and applies low-level optimizations. This makes it possible to upgrade the execution capabilities of existing blockchains without fragmenting the developer experience.
Parallelism at the Opcode Level
A cornerstone of the Altius design is parallel execution, not just across transactions, but within them. Unlike conventional systems that fall back to serial execution on conflicts, Altius analyzes dependencies at the opcode level.
Only the conflicting instruction sets are re-executed, significantly reducing unnecessary computation and allowing finer-grained parallelism than what current L1s and L2s offer.
Sharded State and Lightweight Nodes
Altius pairs its parallel engine with a sharded state layer. This architecture distributes read/write operations across multiple nodes, avoiding the bottlenecks of a single-state database.
This design allows the system to run on commodity hardware, preserving decentralization and accessibility. There is no reliance on specialized validators or centralized sequencers, a clear break from current “high-performance” solutions that trade off decentralization for speed.
A Shared Execution Layer for Multiple VMs
Altius is already EVM-compatible and is actively expanding support for SVM and MoveVM. The broader vision is to enable a shared execution layer that can dynamically route and execute workloads across chains, a unified layer for heterogeneous VMs.
This allows developers to retain their existing smart contract environments while gaining performance gains typically reserved for more centralized chains.
Open Execution Network: Scaling Across Chains
The long-term vision includes the Open Execution Network (OEN), a decentralized mesh of nodes that can dynamically allocate execution capacity. Instead of each chain overprovisioning for peak load, chains could borrow unused capacity from others.
This brings elasticity and efficiency to the execution layer - something that’s missing in the current generation of modular blockchains.
Team and Backing
The team behind Altius was built for this challenge. Annabelle Huang brings experience from Amber Group and early DEX development at AirSwap. Her co-founder and CTO, Anit Chakraborty, spent over 14 years building low-latency systems at Hudson River Trading.
This technical depth attracted early conviction: Altius raised $11M in pre-seed funding from investors including Pantera Capital, Founders Fund, and Amber Group. Not for a narrative, but for the depth of execution-layer engineering.
Roadmap and What's Next
Altius is rolling out its EVM implementation in beta this quarter, with early design partners already building on it. Support for SVM and MoveVM is on the roadmap, alongside longer-term development of the Open Execution Network.
The project is still early, but the architecture is positioned to redefine execution in a multichain world, not by adding another chain, but by upgrading how they all work.