Bodo and Xilinx Collaborate To Bring Python Simplicity to Large-Scale Media Processing

READ MORE

When To Use Bodo


The Bodo platform powers fast, efficient big data processing for Python data teams. Bodo’s Inferential Compiler delivers high-performance style computing for data-intensive processing. Bodo analyzes the syntax of regular Python code to determine opportunities for parallelization and infers an optimal code structure to generate parallelized binary code. The Bodo platform also includes: High-performance connectors, parallel I/O, SaaS notebooks facilities, and resource management within the cloud infrastructure.

A Shining Star Icon

Where Bodo Shines

The clearest speed and efficiency improvements come with using Bodo with other supported Python libraries, as well as with large data processing intensive use cases that benefit from parallelized execution.

Ideal Data Characteristics and Use Cases:

Bodo’s linear scaling capability is most noticeable with efforts involving jobs of 100’s of GBs, hundreds of millions of Dataframe rows, and compute times approaching/exceeding 1h. Bodo best addresses the following use cases and pain points.

Pain Points


  • Long Processing Time
  • Time to Translate to Better Performing Languages
  • Lack of Parallel Programming Skills
  • Lost Time by Analytics Team Awaiting Data
  • Long Processing Time

    Long jobs -- taking hours or days instead of minutes or seconds, causing missed SLAs.
  • Time to Translate to Better Performing Languages

    Data teams lose time moving from prototyping to deployment when code needs to be translated to a different language to achieve better performance.
  • Lack of Parallel Programming Skills

    Many Data Engineering staff lack parallel programming skills in Scala, C, Spark, etc.
  • Lost Time by Analytics Team Awaiting Data

    When analytics teams sit idle, awaiting data from data prep teams.

Use Cases


  • Data Prep and ETL
  • Ml Model Training
  • Feature Engineering
  • Exploratory Analysis
  • Data Prep and ETL

    Data transformation to integrate and format data to be ready for analysis, reporting, and machine learning.
  • Ml Model Training

    Where fast, efficient ingestion and transformation of large data sets is needed.
  • Feature Engineering

    Data analysis to reveal categories, properties, and attributes of data.
  • Exploratory Analysis

    For big data that includes large and/or repetitive data set analyses.

Optimized Python Libraries:

Bodo compiles functions into efficient native parallel binaries, which require that the operations used in the code are optimized by Bodo. This excludes some Python features. Optimized libraries include:


  • Pandas
  • Numpy
  • Machine Learning
  • Deep Learning
Greater than or Equal To Icon

Minimum Requirements:

Basic CPUs (e.g., on-premises, AWS, Google Cloud, Azure). Bodo does not require any special-purpose hardware or networking.