About US
CheetahDS is a spinoff from the University of South Florida (USF), Tampa, Florida. The CheetahDB system is built based on rigorous database system research in the group of USF computer science professor Yicheng Tu. The group has worked on GPU computing since 2009, under the support of multiple National Science Foundation (NSF) and National Institutes of Health (NIH) grants and was named a CUDA Teaching Center and CUDA Research Center by NVidia.
Our vision for the next 5 years is to become a global leader in the high-performance database software market. We will also develop application programs on top of CheetahDB for key customer segments and become a strong player in the OLAP and data stream processing market.
Our vision for the next 5 years is to become a global leader in the high-performance database software market. We will also develop application programs on top of CheetahDB for key customer segments and become a strong player in the OLAP and data stream processing market.
Our Technology and Its Advantage
CheetahDB is based on a complete rethinking of the software architecture of a DBMS under today’s big data application.
GPUs - A Novel Platform for Running Database Queries• Thousands of Computing Cores
• High Memory Bandwidth (up to 900GB/s) • Low Cost ($1000-$2000) • Low Energy Consumption (e.g., 250W) |
GPU-based Database Management System (G-DBMS)• Software development on GPUs is tricky
• Complete rethinking of DBMS architecture • Existing GPU databases: insufficient performance • Basic CS research —> Successful software engineering |
Current and pending patent applications related to the CheetahDB technology
The above patents are based on recently published research that delivered superior performance in key aspects of database systems such as join algorithms, user-defined functions, multi-query optimization, and data indexing. We are confident that we are the technical leaders of this field.