Proposal: Sunsetting Phala parachain and Migrating to Ethereum L2

This proposal has also been published as a on-chain referendum, please come and vote for it:
Vote here, Proposal 8

TL;DR

Phala’s parachain slot will expire on November 20, 2025. The network has remained stable with around 65M PHA delegated and nearly 29,000 miners active, continuing to operate on Intel SGX, which has provided a reliable foundation for Phala’s confidential computing to date. Looking ahead, Intel’s confidential-computing roadmap is centered on Intel TDX and complementary GPU-based solutions, which are designed to extend the capabilities pioneered by SGX and to better serve the scalability and requirements of future enterprise adoption.

This proposal recommends sunsetting the Phala parachain and migrating to Ethereum L2, consolidating staking, governance, and confidential compute in a stronger, EVM-aligned ecosystem.

Background

Over the past year, the Phala team has shifted much of its effort to the next generation of confidential computing, focusing on Intel TDX CPUs and GPU Confidential Compute. These technologies are where Intel and the broader industry are investing, and they align closely with the needs we see from enterprise users. We have already conducted early commercial trials in these areas, confirming their potential.

By contrast, the Phala parachain continues to run on Intel SGX, a technology that was groundbreaking when introduced and still provides a stable foundation today. However, Intel has since shifted its roadmap toward Intel TDX and GPU-based confidential computing, which are designed to address broader enterprise requirements and future scalability. These newer technologies build upon the foundation SGX established, but offer a more sustainable path for long-term ecosystem growth and commercial adoption.

Meanwhile, the parachain slot itself is approaching expiration. Keeping it alive would consume significant resources while locking us into an infrastructure with limited scalability. Migrating to L2 instead allows us to:

  • Focus our resources on TDX and GPU compute, where the future lies.
  • Align with the EVM ecosystem, where developer activity, liquidity, and tooling are strongest.
  • Provide a scalable foundation for staking, governance, and confidential compute workloads.

Thus, we propose to sunset the Phala parachain and fully migrate to Ethereum L2.

Current Status

Phala Parachain

  • The Phala parachain slot is expected to expire on November 20, 2025.
  • Since the sunsetting of Khala, Phala Network delegation has remained stable at around 65 million PHA, with approximately 29,000 miners consistently running.
  • The network runs on Intel SGX, which Intel now maintains only for existing deployments. Future Xeon platforms focus on TDX, leaving SGX with declining hardware and ecosystem support.
  • Continuing to operate SGX-based miners would require ongoing investment into a technology with limited long-term viability.

Phala L2

In contrast, Phala L2 is already live and functional:

  • Explorer: explorer.phala.network provides transparency of chain activity.
  • Bridge: bridge.phala.network enables seamless PHA transfers between Ethereum and L2.
  • KMS Contract: A core component of Phala Cloud security infrastructure is already deployed, proving the readiness of confidential compute on L2.

L2 offers lower operational overhead and direct integration with Ethereum liquidity and tooling. It is also the natural home for deploying TDX and GPU-based confidential compute workloads, where we already see early-stage commercial traction.

Moving forward, Phala L2 will serve as the execution layer for GPU and TDX confidential compute, while staking and governance remain secured on Ethereum L1. This dual-layer design ensures continuity for token holders and delegators, while enabling the ecosystem to scale into next-generation confidential compute.

Migration Plan

If the proposal passes, the following steps will be executed:

Phase 1 – Community Decision & Communication

Sep 26th - Oct 15th

First, the community will vote on this proposal to decide whether to proceed with the sunset of the Phala parachain.

If approved, the team will immediately publish a detailed announcement, including a timeline, FAQs, and user guides to ensure everyone understands the process.

Phase 2 – Development & Exchange Preparation

Oct 15th - Nov 20th

The team will complete all necessary development work to ensure a smooth migration:

  • Finalizing the Ethereum L1 staking contract and vPHA issuance logic.
  • Building the claim interface in the Phala App for asset migration.
  • Conducting audits and internal testing to ensure contracts and tools are safe and reliable.

In parallel, we will work with exchanges (Gate.io) to gradually wind down phala parachain deposit/withdrawal support and prepare for L2 integration.

This phase ensures that when the parachain shuts down, all required infrastructure (staking, claiming) and exchange transitions are already in place.

Phase 3– Phala sunset execution

Nov 20th

Upon slot expiration, we will proceed with the shutdown of collators, Gatekeepers, and other parachain infrastructure.

During the execution period, all assets on the Phala parachain will be frozen and cannot be transferred. They will remain safely locked until the asset claim process is opened in Phase 4, at which point users will be able to redeem them into Ethereum L1/L2 through the Phala App.

Users will be informed well in advance of the exact timeline for this shutdown.

Phase 4 – L2 Activation & Asset Migration

Nov 20th - Nov 25th

Phala L2 will officially become the new home for staking, governance, and confidential compute.

Staking will continue through the Ethereum L1 contract, where PHA can be staked and converted into vPHA. vPHA will serve as the functional token within the ecosystem, used for community governance, miner rewards, and other protocol-level utilities.

For asset migration, the Phala App will provide a one-stop interface:

  • Users can claim their parachain assets directly into L2-connected addresses.
  • Existing delegations will be mapped into the new vPHA staking system.
  • Any previously unclaimed Khala assets will also be recoverable through the same process.

Phase 5 – Governance

Nov 26th - Nov 30th

Once the migration is complete, governance and treasury management will be conducted on Ethereum L1, using Snapshot + Safe multisig as the primary mechanisms. vPHA will serve as the governance utility token, ensuring that staking participants and miners can fully exercise their voting rights and participate in community decision-making.

Phase 6 – Long-Term Roadmap

Long term plan

On the mining side, participants will run GPU workloads on L2, and their rewards will also be distributed in vPHA. The exchange rate between vPHA and PHA will be automatically adjusted by the L1 staking contract as rewards are distributed, ensuring that users face no asset risk and that all vPHA can always be redeemed back into PHA.

The team will also focus on developing a usage-based settlement model for TDX + GPU compute, and positioning PHA as the core settlement token across these services.

Impact on Users & Ecosystem

  • PHA Token Holders: Full asset claimability, uninterrupted staking opportunities.
  • Delegators: Existing delegations will transition smoothly into the new vPHA staking framework, without interruption.
  • Miners: While SGX miners face structural limitations, the migration enables participation in GPU/TDX mining on L2, with rewards distributed in vPHA that can always be redeemed back to PHA.
  • Developers/Partners: Stronger Ethereum/EVM integrations, liquidity, and tooling.
  • Community: Governance and treasury management will operate on L1 via Snapshot + Safe, with vPHA as the governance utility, ensuring decentralization and transparency.

This is not the end of Phala. It’s the beginning of a new chapter, where we embrace Ethereum L2 and the next generation of confidential compute — powered by TDX CPUs and GPU acceleration.

This doesn’t make much sense, from a security perspective. I think that if you want to build something which is side-channel resistant you should not allow regular users to host enclaves. These should be cloud offerings where there’s no physical access, or if there is, it’s highly regulated.