VectorSphere
VectorSphere is a next-generation HTAP DBMS providing real-time processing and large-scale analytics. Featuring a cloud-native architecture, vectorized execution engine, and CDC-based real-time data integration, it is the optimal solution for the digital transformation era.
Development Overview
- Background : Next-generation DW solution developed since 2021, led by core R&D members formerly from Greenplum.
- Proven Track Record : Secured over 200 operational enterprise clients across various industries, including finance, manufacturing, telecommunications, and automotive.
- Partnerships : Proven stability and performance through collaborations with global cloud providers and major SI partners.
Key Challenges Faced by Customers
Storage Engine Confusion
Increased complexity and management overhead due to mixed formats such as Heap, AO, and AOCO.
Lack of Real-Time Processing
Inability to handle T+0 real-time processing due to a focus on T+1 workloads.
Performance Instability
Performance degradation caused by catalog bloat and delayed statistics updates.
Absence of CDC Solution
Complete lack of real-time data integration and analytical capabilities.
Innovative Tech Stack
- Mars3 Storage Engine : A unified, single-engine, column-based compression/sorting system. Supports cloud-native architecture via S3 integration.
- MxVector Execution Engine : A high-performance execution engine that delivers 10x to 100x query performance boosts through SIMD-based vectorized processing.
- Domino Stream Engine : A streaming processing engine supporting real-time data transformation, Continuous Views, and CDC.
Overwhelming Performance Advantage vs. Greenplum 7
- Delivered 10x higher TPC-H and 100x higher SSB performance compared to Greenplum 7, as well as a 40%+ performance edge over Vertica—proving superior efficiency against competing solutions.