Built from years of HPC research, Bodo is the world’s first auto-parallelizing inferential Python compiler.
Bodo generates low-level, parallel MPI code directly, sidestepping the inefficiencies and complexities of traditional frameworks. This approach delivers true HPC-level performance with minimal effort, enabling Python users to easily leverage the power of HPC for analytics, AI, and more.
We are dedicated to Python and aspire to make Bodo as easy to use as possible – no rewrites or complicated framework integrations.
Bodo delivers HPC-level performance and supports the Python libraries you already know and love, making it easy to scale with Python
Bodo binaries are infrastructure-agnostic, scaling from a single laptop to cloud environments as you add cores—no additional code needed.
Bodo is engineered to bypass the bottlenecks that slow down other frameworks like Spark, Ray, or Dask— delivering 10x to 100x+ faster performance on similar workloads
Bodo maximizes resource utilization, reducing waste and compute expenses. This efficiency also shrinks environmental footprint.
This is the real deal. Bodo built on the success of Numba to combine compiled Pandas and automatic parallelism (with MPI) to get incredibly fast data processing using simple syntax. It can make your code using Python *fast*—simply.
Bodo fits into your existing workflows, including popular libraries like NumPy, SciPy, Pandas, and even SQL. Features include:
Automatic Compiler Parallelization & MPI Code Generation
Comprehensive Pandas & Numpy API Support
Compiler Optimizations for Sequential CPU Performance
Enhanced Support for Data Science & Machine Learning Libraries
Python UDF Compilation & Optimization
Scalable I/O for Leading Data Formats & Databases
Automatic Filter Pushdown
AI Training Integration
Native Python & SQL Integration
Portable Compilation with LLVM
Audit logs, user roles, access controls, and flexible deployment options to fit your needs.
Bring Bodo to whichever environment you choose.
Learn how Bodo enabled on-demand repair for 1.8 billion devices by standardizing and accelerating part forecasting.
Learn more