BabelBirdBabelBird Docs

Xinchuang Support

BabelBird is compatible with ITAI/Xinchuang environments. Both the BabelBird Enterprise Drive server and Zhichao AI private deployment can be planned for Xinchuang infrastructure. The final design should consider existing infrastructure, database licensing, CPU architecture, operating system, middleware, storage and AI compute resources.

General Recommendation

x86 CPUs are recommended first for BabelBird server and Zhichao AI private deployment. The x86 ecosystem usually has more mature software compatibility, driver support, model inference toolchains and operational experience, making it suitable for most production environments.

BabelBird is also compatible with ARM environments and can be adapted for Kunpeng, Feiteng ARM and similar server architectures. Before choosing ARM, evaluate compatibility for databases, middleware, search components, object storage, AI inference frameworks and third-party drivers.

Server Support

The BabelBird server can run in environments that support containerized or Linux service deployment. Common Xinchuang operating systems include openEuler, Kylin and UOS. Deployment mode can be stand-alone, clustered, object-storage based or hybrid depending on scale.

Layer Standard deployment Xinchuang x86 option Xinchuang ARM option Notes
CPU architecture x86 Hygon, Zhaoxin and other x86 Xinchuang servers Kunpeng, Feiteng ARM and similar Evaluate x86 first for production
Operating system CentOS, Ubuntu, openEuler openEuler, Kylin, UOS ARM editions of openEuler, Kylin, UOS Depends on container, kernel and driver compatibility
Application service Tomcat and similar Tomcat, TongWeb and similar Tomcat or adapted middleware Third-party middleware licensing must be confirmed
Search component Elasticsearch Xinchuang-compatible edition ARM-compatible edition Used for full-text search, OCR indexing and log search
Cache Redis Redis or compatible components Redis/ARM-compatible components Confirm high-availability design
Object storage MinIO or third-party object storage MinIO, domestic object storage or enterprise storage MinIO ARM or compatible object storage Should follow private storage architecture

Database Support

In Xinchuang environments, BabelBird supports MySQL-compatible databases by default. GreatSQL, GaussDB for MySQL, AliSQL, TiDB and other MySQL-compatible solutions are recommended for priority evaluation. Database selection should focus on transactions, character sets, indexes, connection pools, backup/restore, replication or clustering, high availability and operations tooling.

BabelBird can also integrate with Dameng, Kingbase, GBase and other domestic databases. These databases usually require compatibility development, SQL adaptation, performance validation and regression testing. Enterprises also need to purchase the database license separately and cover the database compatibility development cost.

Zhichao AI Private Deployment

Zhichao AI private deployment can also be planned for Xinchuang environments. Compared with the regular drive server, AI deployment must additionally evaluate model runtime, GPU/AI accelerator cards, drivers, inference frameworks, vector retrieval, OCR, text-to-image search, image-to-image search and knowledge-base indexing.

Confirm before implementation:

  1. Whether to use traditional OCR or AI OCR.
  2. Whether private large models, embedding models, multimodal models and knowledge-base robots will be deployed.
  3. Whether GPU or AI accelerator cards have drivers and inference framework support for the selected Xinchuang system.
  4. Whether model services and drive business services should be deployed separately.
  5. Whether AI indexes, caches, logs and audit data meet enterprise compliance requirements.

Client Support

BabelBird provides Linux/Kylin client capabilities for certain domestic desktop environments. Because Linux desktop file-manager interfaces are not unified, the Linux client currently does not support folder synchronization or mapped drives. If synchronized folders, virtual mapped drives or deep local file-manager integration are required, Windows or macOS clients are recommended first, or custom development should be evaluated separately.

Selection And Implementation Guidance

  • Confirm required CPU, operating system, database and middleware lists first.
  • For MySQL-compatible databases, test connection, transactions, indexes, character sets, backup/restore and performance.
  • For non-MySQL-compatible databases, reserve adaptation development and regression testing time.
  • For Zhichao AI private deployment, evaluate models, GPU, OCR, multimodal and vector retrieval resources separately.
  • Prepare disaster recovery, backup, logs, monitoring, upgrade and rollback plans for production.
  • Confirm in advance whether clients need synchronized folders, mapped drives and local editing integration.
BabelBird capabilities may change by product version, licensed modules and deployment configuration; actual availability depends on the deployed environment and administrator settings.