SOON Network strategic investment in the Phala TEE GPU cluster, providing privacy-preserving and verifiable computing power for AI Agents 🚀
SOON’s core infrastructure development team has become a key funding backer of the PhalaNetwork TEE GPU cluster. In the future, AI Agents in the SOON ecosystem will be able to directly call this compute power to run customized models in a fully private, trusted environment.
Why is this so important?
AI Agents entering capital markets need not only extreme execution performance, but also privacy-preserving and verifiable computing support. SOON has now perfectly combined both:
• Full strategic privacy, execution is verifiable on-chain: For quantized Agents, customized models and trading strategies are their core assets. Running inside Phala’s TEE ensures that models and data remain encrypted end to end; meanwhile, remote attestation allows anyone to verify that each execution strictly follows the expected behavior.
• Supports any team to deploy dedicated SVM chains: SOON Stack is currently the only SVM solution that can deploy a dedicated chain for third parties with one click on multiple mainstream L1s, such as Ethereum, BNB Chain, and Base. Each chain benefits from ultra-high throughput and ultra-low latency enabled by parallel execution.
Our vision
So that every Agent can run a dedicated chain with “Solana-level performance”: self-manage block space and compute fee payments, enabling 24/7 unattended operation, while the owner always retains ultimate control. SOON AI capital markets will serve as the core execution and settlement layer for these Agents.
Next steps
• The compute revenue sharing mechanism and how it links with the SOON ecosystem will be disclosed in a subsequent announcement;
• The onboarding plan for the first wave of AI capital market projects is about to be announced.
“AI capital markets need more than speed—they need privacy-preserving and verifiable computing—this is the core enablement Phala brings to $SOON .”
—— Marvin Tong, Phala Network
Note: This announcement contains forward-looking statements, and subsequent plans may be adjusted based on actual circumstances.
SOON’s core infrastructure development team has become a key funding backer of the PhalaNetwork TEE GPU cluster. In the future, AI Agents in the SOON ecosystem will be able to directly call this compute power to run customized models in a fully private, trusted environment.
Why is this so important?
AI Agents entering capital markets need not only extreme execution performance, but also privacy-preserving and verifiable computing support. SOON has now perfectly combined both:
• Full strategic privacy, execution is verifiable on-chain: For quantized Agents, customized models and trading strategies are their core assets. Running inside Phala’s TEE ensures that models and data remain encrypted end to end; meanwhile, remote attestation allows anyone to verify that each execution strictly follows the expected behavior.
• Supports any team to deploy dedicated SVM chains: SOON Stack is currently the only SVM solution that can deploy a dedicated chain for third parties with one click on multiple mainstream L1s, such as Ethereum, BNB Chain, and Base. Each chain benefits from ultra-high throughput and ultra-low latency enabled by parallel execution.
Our vision
So that every Agent can run a dedicated chain with “Solana-level performance”: self-manage block space and compute fee payments, enabling 24/7 unattended operation, while the owner always retains ultimate control. SOON AI capital markets will serve as the core execution and settlement layer for these Agents.
Next steps
• The compute revenue sharing mechanism and how it links with the SOON ecosystem will be disclosed in a subsequent announcement;
• The onboarding plan for the first wave of AI capital market projects is about to be announced.
“AI capital markets need more than speed—they need privacy-preserving and verifiable computing—this is the core enablement Phala brings to $SOON .”
—— Marvin Tong, Phala Network
Note: This announcement contains forward-looking statements, and subsequent plans may be adjusted based on actual circumstances.
