More technical exchanges, job search opportunities, welcome to pay attention to theByteDance Data PlatformWeChat Public，Reply [1]Enter the official communication group.
The new energy vehicle market is ushering in a period of rapid development. According to IDC forecast, the market size of new energy vehicles in China's passenger car market will exceed 23 million units in 2028, with a compound annual growth rate of 22.8%.
A set of highly reliable, high-performance and highly available data analysis system is of great significance for new energy vehicles to find and solve problems in time, ensure vehicle safety and improve product quality.
There have been news reports in the industry of high battery temperatures that exceeded safety thresholds and led to vehicle accidents. The Real-time Vehicle Signal Data Analysis System, on the other hand, monitors battery temperature, current, voltage and other signal data in real time. When the temperature rises abnormally, the system is able to send an immediate alert to the vehicle owner, reminding him or her to take measures, such as reducing the speed of the vehicle or finding a safe place to park as soon as possible. Similarly, the relevant data is transmitted back to the vehicle manufacturer's servers in real time. The manufacturer's technical team can quickly analyze the data to determine whether the problem is an individual vehicle or a batch product quality issue.
In order to support the real-time requirements of the vehicle data system, car companies tend to favor analytical databases that can analyze large-scale data and complex scenarios in the selection of the underlying data engine. As an analytic database for OLAP launched by Volcano Engine, ByteHouse has entered the field of vision of a certain series of energy vehicle enterprises due to its high performance and extreme analytic capability.
By selecting one day's sample data of a certain car, the car company simulated nearly 100 billion pieces of data for testing. In query scenarios such as single-table point lookup, single-table aggregation, and correlation aggregation, based on the same SQL query, ByteHouse performance improves at least 4 times compared to the same type of products in the market.
According to the introduction, ByteHouse's high performance mainly comes from its series of optimization measures in complex queries, wide table queries and other scenarios. On complex queries, ByteHouse has launched a series of self-developed optimizers, including RBO (Rule-Based Optimization), CBO (Cost-Based Optimization), Distributed Plan Generation, etc., which can accurately compute the efficiency maximization execution path and significantly reduce user query time. In addition, ByteHouse is also optimized in the directions of Exchange, Runtime Filter, and parallelization refactoring. In the wide-table query scenario, ByteHouse mainly uses Global Dictionary, Zero copy and UncompressedCache to achieve performance improvement.
In a previously released performance whitepaper, ByteHouse demonstrated its hardcore performance through the results of SSB, TPC-H and TPC-DS dataset tests. Using an open source OLAP known for its performance as a benchmarking product, ByteHouse shows significant performance improvement in different query items. Taking the TPC-H dataset as an example, ByteHouse's query efficiency is dozens of times higher than that of the benchmark product under the same hardware and software environment.
Through a series of technical optimization means, ByteHouse achieves further performance enhancement, shortens query execution time, optimizes resource utilization, can cope with more complex query scenarios, and provides users with a smoother data analysis experience, which is applied in the fields of Internet, game, finance, automobile, meteorology, and so on, and helps to promote the transformation and upgrading of digital intelligence.
click to jumpVolcano Engine Cloud Native Digital Warehouse ByteHouse Learn more.