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


Simpler, Faster
Parallel Computing.

Bodo is a new approach to HPC-style parallel computing, but with the simplicity of native Python. Our compiler technology infers the parallel structure of application code, including Pandas, NumPy and others, and generates true parallel execution code - with speed and linear scalability beyond 10,000 cores

> conda create -n Bodo python=3.9 -c conda-forge
> conda activate Bodo
> conda install bodo -c -c conda-forge
def data_transform():
  df = pd.read_parquet(input_parquet_file)
  df["B"] = df.apply(lambda r: "P1" if r.A.month < 5 else "P2", axis=1)
  df["C"] = df.A.dt.month
Use Native Python

Use Native Python;
No Rewrites to Scale

Bodo enables faster development and easier maintenance for data science, data engineering and ML workloads with standard Python APIs like Pandas.

Faster Than Spark, Ray or Dask

10x-100x Faster Than
Spark, Ray or Dask

Bodo’s powerful compute engine and parallel computing approach provides efficient execution and effective scalability even for 10,000+ cores and PBs of data.

Predictable, Efficient and Cost-Saving

Predictable, Efficient and 90% Cost-Saving

Highly linear scaling and CPU utilization drives up to 20x increase in hardware efficiency.

Bodo logo
Extreme Performance Analytics Platform
Higher Performance ConnectorsAutomatic ParallelizationEnd-to-End OptimizationHigh Performance Analytics PrimitivesIntegration Tools
*Coming Soon

Use Bodo When and Where You Want

Deploy On-Premises or Your Preferred Cloud

Deploy On-Premises or
Your Preferred Cloud

Bodo’s engine can be used on your laptop, with containers, and with any popular cloud platform.

Data Warehouse

Speed Processing on Your Existing Data Warehouse or Data Lake

Bodo can be used with any data source, including popular cloud-based data warehouses and LakeHouses.

Use With Popular Data Tools and APIs

Use With Popular Data Tools and APIs

Save development and learning curves. Use familiar Pandas, SciKit-Learn, NumPy, and more.

Bodo Overview

Watch a two minute summary of how Bodo works, and how it simplifies work for Data Engineers and Data Scientists.

Clear Business Impact

Efficient use of CPU

Save up to 90% on Infrastructure Costs

Efficient use of CPU cores, and elimination of wait states means less overall CPU time, and less corresponding cost.

Reduce Time-to-Business-Value

Reduce Time-to-Business-Value

Shorter developer learning curves vs. existing frameworks and libraries; eliminate need to hand-off tasks between programming teams.

architecturally highly resilient

No More Failed Jobs.
No More Wasted Time.

Bodo is architecturally highly resilient to software and memory overload failures during runtime. Especially valuable when running large jobs.

Try Bodo On Your Terms...For Free

bodo logo
Enterprise Edition

Extreme performance at massive scale but with lower cost to help you achieve your business goals.

bodo logo
Cloud Platform

Launch Bodo-powered clusters, jobs and notebooks within minutes on a fully managed cloud platform.

bodo logo
Community Edition

Accelerate your Python data science code automatically on up to four cores with the free Bodo Community Edition.

Get Connected

Join our Slack, Feedback Github, and Discourse.