scale-out deterministic execution

This work addresses the performance bottleneck in blockchain execution layers caused by the inability of single-node validators to meet the growing demands of smart contract execution. To overcome this limitation, the authors propose a horizontally scalable, deterministic execution engine featuring an asymmetric architecture that combines centralized scheduling with distributed execution. The design decouples consensus from execution through a stateless-stateful separation mechanism and enforces execution determinism via a strict ownership model based on object versioning. Furthermore, a locality-aware and load-balanced scheduling strategy is introduced to elastically adapt to bursty workloads and dynamic access patterns. Experimental results demonstrate that the system achieves up to 250,000 transactions per second—three times higher throughput than existing solutions—with latency reductions of up to 5 milliseconds, aligning execution performance with modern consensus protocols.

Ethereum faces scalability bottlenecks due to the EVM’s inherently serial execution model. This paper proposes a native parallel execution framework for the EVM that overcomes sequential constraints. First, it introduces a state-access predictability framework, leveraging static analysis and state-dependency graph modeling to proactively detect read–write conflicts across transactions. Second, it designs a gas-driven parallel incentive mechanism that guides efficient, concurrency-safe transaction scheduling. Third, it implements a lightweight, bytecode-level modification to the EVM, preserving full backward compatibility. Experimental evaluation demonstrates that the approach achieves a 3.2× improvement in throughput (TPS) and reduces block confirmation latency by 57%, while maintaining strict EVM semantic equivalence. The results indicate near-linear scalability potential under realistic workloads.

To address concurrency control challenges under mixed long/short transaction workloads in manufacturing systems, this paper proposes a decentralized graph-based protocol that guarantees zero aborts for long update transactions, improves short-transaction throughput, and fully exploits multicore parallelism. Our key contributions are: (1) the first decentralized scheduling mechanism based on a full-precision multiversion serialization graph (MVSG), enabling lock-free concurrency and distributed graph maintenance; and (2) BoMB—the first OLTP benchmark tailored to Bill-of-Materials (BOM) scenarios—accurately modeling heterogeneous transaction conflict patterns. Experimental evaluation on BoMB demonstrates 100% commit rate for long transactions, short-transaction throughput of 1.7 Mtpm, and near-linear scalability—substantially outperforming state-of-the-art approaches.

This work addresses performance bottlenecks—low throughput, high tail latency, and poor scalability—of distributed locks under high contention and geo-distributed active-active deployments. We propose a latency-aware distributed locking mechanism: leveraging multi-datacenter topology modeling, it integrates a lightweight consensus protocol design with latency-aware scheduling, achieving substantial reduction in cross-region coordination overhead while preserving strong consistency. Experimental results demonstrate a 68% throughput improvement and a 52% reduction in P99 latency under high contention, with near-linear scalability as the number of nodes increases. Unlike conventional centralized or classical distributed lock schemes, our approach is the first to achieve both high efficiency and scalability in geo-distributed active-active settings while guaranteeing linearizability. The proposed lock primitive offers practical deployability, flexibility across heterogeneous infrastructures, and serves as a next-generation foundation for large-scale distributed systems.

This work addresses the challenge of efficiently supporting serializable transactions in geo-distributed databases under high network latency, where excessive coordination severely limits performance. The authors propose Minerva, a system that decouples data propagation from transaction commit through epoch-based asynchronous replication and combines optimistic concurrency control with deterministic re-execution to enable high-throughput, coordination-free, multi-master serializable transactions. Its key innovation lies in integrating epoch-based replication with deterministic re-execution, modeling conflicts via a conflict graph and optimizing transaction commits using a maximum-weight independent set algorithm to substantially reduce re-execution overhead. Experimental results on the TPC-C benchmark demonstrate that Minerva achieves over 3× higher throughput than existing systems under typical conditions and up to 2.8× improvement in high-latency scenarios.

Current DevOps infrastructures for blockchain applications are predominantly controlled by single entities, lacking decentralized deployment and governance mechanisms. This work proposes a decentralized deployment architecture decoupled from specific governance and upgrade schemes, integrating DAO-based governance, smart contract upgradability, and DevOps best practices. By adopting an extended registry pattern, the architecture enables deterministic deployments and, for the first time, incorporates version control, testing and validation, and user interface components into a unified decentralized framework. The project provides an open-source reference implementation that substantially lowers the barrier to practical decentralized deployment. Experimental evaluation demonstrates the effectiveness and practicality of the proposed architecture.

This work addresses the correctness challenges in implementing linearizable atomic registers in asynchronous message-passing systems, where precise real-time ordering of operations is unavailable. By combining equivalence and indistinguishability arguments with message-chain theory, the paper rigorously establishes that ensuring linearizability necessitates the formation of extensive message chains between operations of any type. This result formally characterizes, for the first time, the inherent communication overhead imposed by linearizability in asynchronous settings, thereby establishing a fundamental lower bound on the communication complexity required for its implementation. The findings provide a theoretical foundation for understanding the structural constraints and design costs associated with achieving linearizable semantics in distributed systems.

This work addresses the challenge of ensuring cross-chain atomicity under asynchronous communication and Byzantine node settings by proposing a cross-chain protocol that supports composable atomic transactions. The protocol employs a shared coordination layer—comprising a sequencer, transaction processor, coordinator, and confirmation layer—to guarantee all-or-nothing execution of cross-chain operations while preserving individual chain autonomy. It innovatively separates accepted and deferred transaction sets and integrates timeout mechanisms with dependency depth limits, achieving minimal blocking and bounded latency under strong safety and liveness guarantees. Formal verification and experimental evaluation demonstrate that the protocol attains high transaction success rates under moderate cross-chain loads and quantifies the trade-off between success rate and dependency-induced delay.

Traditional distributed systems struggle to support modern autonomous infrastructures that integrate stochastic models and autonomous agents. This work proposes the Post-Deterministic Distributed System (PDDS) model, introducing for the first time its five architectural pillars. Its core innovation is a "cognitive state replication" mechanism that extends consistency from data visibility to knowledge visibility, alongside a novel fault classification framework. By leveraging protocol-driven development, verifiable agent infrastructure, and semantic quorum guarantees, PDDS enables coordination among semantically equivalent yet executionally divergent agents. This approach achieves verifiable semantic rollback and cross-agent reasoning consistency, establishing a theoretical foundation for trustworthy autonomous systems.