Optimistic TEE-Rollups: A Hybrid Architecture for Scalable and Verifiable Generative AI Inference on Blockchain

TL;DR

This paper presents a hybrid architecture called Optimistic TEE-Rollups (OTR) that enables scalable, verifiable, and low-latency generative AI inference on blockchain by combining trusted hardware, optimistic fraud proofs, and stochastic zero-knowledge checks.

Contribution

The paper introduces OTR, a novel hybrid verification protocol leveraging TEEs and cryptographic proofs to address the verifiability trilemma in decentralized AI inference.

Findings

Achieves 99% throughput of centralized systems

Maintains low cost of $0.07 per query

Ensures Byzantine fault tolerance against rational adversaries

Abstract

The rapid integration of Large Language Models (LLMs) into decentralized physical infrastructure networks (DePIN) is currently bottlenecked by the Verifiability Trilemma, which posits that a decentralized inference system cannot simultaneously achieve high computational integrity, low latency, and low cost. Existing cryptographic solutions, such as Zero-Knowledge Machine Learning (ZKML), suffer from superlinear proving overheads (O(k NlogN)) that render them infeasible for billionparameter models. Conversely, optimistic approaches (opML) impose prohibitive dispute windows, preventing real-time interactivity, while recent "Proof of Quality" (PoQ) paradigms sacrifice cryptographic integrity for subjective semantic evaluation, leaving networks vulnerable to model downgrade attacks and reward hacking. In this paper, we introduce Optimistic TEE-Rollups (OTR), a hybrid verification protocol that harmonizes these constraints. OTR leverages NVIDIA H100 Confidential Computing Trusted Execution Environments (TEEs) to provide sub-second Provisional Finality, underpinned by an optimistic fraud-proof mechanism and stochastic Zero-Knowledge spot-checks to mitigate hardware side-channel risks. We formally define Proof of Efficient Attribution (PoEA), a consensus mechanism that cryptographically binds execution traces to hardware attestations, thereby guaranteeing model authenticity. Extensive simulations demonstrate that OTR achieves 99% of the throughput of centralized baselines with a marginal cost overhead of $0.07 per query, maintaining Byzantine fault tolerance against rational adversaries even in the presence of transient hardware vulnerabilities.