The goal of SQuadron is to provide terabyte scale database processing at an affordable price. SQuadron uses cutting edge software techniques and the parallel power of GPU processing to manage and query data efficiently. By utilizing affordable commodity hardware, SQuadron provides cost-effective solutions in support of Big Data analytics.
Multicore CPUs are not sufficient to scale to billions of records. CPUs were designed for complex branch prediction and pipeline processing, and they accomplish this task well. However, this type of processing is not necessary when mining database entries.
GPUs have thousands of cores that are specifically designed to perform remedial redundant processing. This type of processing is perfectly suited for database querying, which can involve processing over billions of records.
Technica was inspired by GPU processing and believes combining these results with other novel software design approaches lays out the path for efficient database processing that can be performed affordably using commodity hardware.
- Dynamic Compression
- Data Skipping
- GPU Acceleration
- Time Series Data
- In-Database Analytics