HyperBEAM brings exotic compute onchain for the first time in history. Building on HyperBEAM is like discovering Arweave before SmartWeave’s inception: from this point, there’s only exponential growth.

This is the HyperBEAM Accelerationism manifesto (hb/acc) – kicking off the initiative through meta-VM and alien compute development led by the Load Network team (Decent Land Labs).

Founded in January 2021, Decent Land Labs' mission has been to innovate by leveraging Arweave’s Permaweb and achieve multichain web3 interoperability, web3 tech abstraction and max UX simplification. This motive led to Load Network back in December 2023 (previously known as WeaveVM).

The main idea driving Load Network was a will to make Arweave’s static storage EVM-computable, hyperscalable on the EVM side of data-centricism, data bandwidth, chain’s data capacity, long-term data availability and bridging the storage interfaces & computation gaps between Arweave and the EVM space.

Load Network development was going alongside ao’s inception and the grand adoption of ao’s legacynet: ao offered a totally new compute paradigm, and Load focused on an EVM compute layer on data-steroids.

At Load, we achieved traction on tech adoption from Dymension’s RaaS integration, EigenDA blobs storage, Ethereum’s Blobscan data storage, L1/L2 ledger archiving + decentralized archive node offering for Metis, RSS3, Avalanche, Humanode, Phala, Rise, GOAT and more. The Load EVM’s technical superiority across the sovereign EVM L1s space has been achieved through our custom work on Ethereum’s Execution client (Reth) work, the novel Arweave interface and data permanency inheritance, data programmability and interoperability.

Manifesto’s reasoning: pragmatism, not romance-idealism

At Decent Land Labs, our building manifesto is driven by pragmatism, and not the romanticization of a tech stack – and this has been demonstrated through our stack choices after careful experimentation and performance alignment, from the first team to adopt Arweave’s 3EM/EXM when it was released, to choosing to custom-build on Reth, and now arriving into the HyperBEAM era.

Our custom work on Reth ExEx being highlighted in devcon DC7

If you speak with developers that know us, they’ll tell you: we build alien stuff and ship fast. And that’s our motto. All that to say – us adopting novel tech such as HyperBEAM isn’t weird. We see the net-new potential in HyperBEAM’s design, its capabilities and what can be built leveraging it. We don't need to stick to a given EVM stack just for the sake of emotional attachment – we can bring whatever we need to use building on EVM to hyperbeam, and that’s the beauty of it. “Be water, my friend”.

What we see with HyperBEAM and ao is unique. We see a golden opportunity to early-adopt a new computing paradigm (ao) and a new decentralized operating system design (HyperBEAM), build on this decentralized supercomputer and leverage the technical superiority of this hybrid stack. Building on HyperBEAM is like discovering Arweave before SmartWeave’s inception: from this point, there’s only exponential growth.

hb/acc TL;DR

The hb/acc manifesto materializes in open source software and build-in-public shipping. HyperBEAM Accelerationism is set to focus on the following angles:

  • Build meta-VMs within the Hyperbeam device stack and achieve interoperability with ao-core compute.
  • Research and build new general purpose utility devices that can result in extra revenue streams for Hyperbeam node operators, beside ao compute. (e.g. GPGPU compute)

We will expand more on the second points in the near future, now focusing on meta-vm acceleration: we want to hb-device-encapulate well-adopted web3 execution machines and make it universally interoperable within each other and ao.

Now imagine that you can go to hyperbeam/ao, and deploy an EVM appchain, or RISC-V EVM appchain, or Starknet’s Cairo-VM one, or Aztec’s Brilling-VM, or even more VMs – this is work in progress and we already have released the MVPs or Revm-based EVM appchain (ultraviolet.load.network) and the first-ever RISC-V execution machine appchain that handle RISC-V smart contracts runtime alongside the EVM.

hb/acc in the Load Network scope

Technical

Load Network will be hb-pilled (it is already actually) and adopt this new technology in the EVM network stack leveraging the devices that will be built along the way; not only that, but also achieve low-level interoperability with ao.

Adopting the hb/acc manifesto, we are set to build all the necessary infrastructure to run an Ethereum (EVM) full node from HyperBEAM, as devices (bytecode execution machine, consensus layer, execution layer, JSON-RPC server, etc) and run Load Network L1 EVM from within the hyperbeam node itself; this includes but not limited to:

  • Bidirectional data and compute pipeline between ao and load network
  • Hybrid EVM gas market: pay EVM gas fee in $LOAD or $AO
  • Asynchronous Precompiles implemented in hyperbeam: read the research blog post
  • EVM blobspace (eip-4844) acceleration: integrate the blobspace with Arweave and ArFleet.

Semi-Technical

On May 30th we announced the inception of Decent Land Labs Research Arm, focused on researching technical and theoretical domains of Arweave, EVM and HyperBEAM/AO. Check the research arm posts here: blog.decent.land.

hb/acc inception in Arweave Day Berlin

On Arweave Day Berlin (June 7th 2025) we announced the inception of the Hyperbeam Accelerationism manifesto, expanding on the points listed in this blog post.

During the talk we demo’d Ultraviolet, the first EVM appchain on top of hyperbeam built on top of the ~evm@1.0 device, and the first general purpose & arbitrary GPU compute, leveraging the ~kem@1.0 device, that’s interoperable with ao atomic NFTs via Bazar.

Final thoughts

hb/acc is a manifesto initiated by us, but not exclusive to any party, it’s an idea to evangelise HyperBEAM and ao adoption, extend its stack and join the ongoing forces of reinforcing the hb/ao dev ecosystem – get hb-pilled & hb/acc!