Research Report on the History of Computer Chinese Chess (Xiangqi) Game-Playing

A structured history of Chinese chess engine development from the 1980s to 2026, covering major engines, protocols, and community tooling. Multi-language Interface Support → Topic 1: Explanation of Core Xiangqi Engine Parameters

☰ Contents

Multi-language Interface Support

Xiangqi engine GUI and interfaces support multi-language display:

Chinese interface: Supports Simplified and Traditional Chinese English interface: Supports full English interface Bilingual display: Simultaneously displays Chinese and international move notation

International Standardization of Game Annotations

Standardization of game annotations facilitates the international dissemination of engine analysis:

Chinese annotations: Use Chinese to describe engine analysis results English annotations: Use English to describe engine analysis results Symbolic annotations: Use universal symbols to describe evaluation values (e.g., +1.50 means Red is ahead by 1.5 pawns)

Volume XXIII: Xiangqi Engine Security Considerations

License Compliance

Licensing compliance issues to note when using Xiangqi engines:

Open source engine licensing: Pikafish uses the GPLv3 license, which must be complied with (provide source code when distributing) Commercial engine licensing: Commercial engines require purchased licenses; they cannot be used, copied, or distributed illegally Data source compliance: The source and license of training data must comply with relevant regulations

Online Platform Usage Standards

Standards for using engines on Xiangqi online platforms:

Gaming platform rules: Most online platforms prohibit engine-assisted play Analysis mode: Using engines in the platform’s analysis mode or spectator mode is permitted Teaching use: Using engines in Xiangqi teaching and training scenarios is legal and recommended

Special Topic 001: Xiangqi Engine Knowledge Points

Topic 1: Explanation of Core Xiangqi Engine Parameters

Xiangqi engines can be configured with multiple core parameters during operation, which have important impacts on engine strength, performance, and stability.

Parameter 1: Hash Table Size

The hash table size determines how many searched position results the engine can cache. The larger the hash table, the more search information the engine can remember, avoiding repeated searches.

Recommended values:

  • 2GB memory system: 64-128MB
  • 4GB memory system: 128-256MB
  • 8GB memory system: 256-512MB
  • 16GB memory system: 512-1024MB

Parameter 2: Number of Threads

The number of threads determines how many CPU cores the engine uses for parallel search. More threads means higher parallel search efficiency.

Recommended values:

  • Dual-core CPU: 2 threads
  • Quad-core CPU: 4 threads
  • Octa-core CPU: 8 threads
  • 16+ core CPU: 8-16 threads

Parameter 3: Eval File Path

The evaluation file path specifies the storage location of the NNUE neural network weight file. The engine loads this file at startup for position evaluation.

Parameter 4: Rule Selection

The rule selection parameter allows users to choose the Xiangqi rule system. Supports two options: Asian rules and Chinese rules.