Scale Python across 1,000's of computers using one line of code.

Burla is a Python package with one function: remote_parallel_map. Here's an example:

This realtime example runs my_function on 1,000 separate computers in one second!

Enable anyone to process terabytes of data in minutes, not days.

Burla is simple enough for anyone to learn, yet extremely scalable, and flexible/configurable.

• Scalable: See our demo where we process 2.4TB in 76s using 10,000 CPUs!

• Flexible: Runs any code, inside any Docker container, on any hardware like GPU's or TPU's.

Easily monitor long-running workloads, or manage compute resources in the dashboard.

How it works:

With Burla, running code in the cloud feels the same as coding on your laptop:

return_values = remote_parallel_map(my_function, my_inputs)

When functions are run with remote_parallel_map:

• Anything they print appears locally (and inside Burla's dashboard).

• Any exceptions are thrown locally.

• Any packages or local modules they use are (very quickly) cloned on remote machines.

• Code starts running in under one second! Even with millions of inputs or thousands of machines.

Features:

📦 Automatic Package Sync

Burla automatically (and very quickly) clones your Python packages in every remote machine where your code runs.

🐋 Custom Containers

Easily run code in any Docker container. Public or private, just paste an image URI in the settings, then hit start!

📂 Network Filesystem

Need to get big data into/out of the cluster? Burla automatically mounts a cloud storage bucket to ./shared in every container.

⚙️ Variable Hardware Per-Function

The func_cpu and func_ram args make it possible to assign big hardware to some functions, and less to others.

Convert any workload into a scalable data-pipeline:

Have a workload that takes forever to run?

By injecting many remote_parallel_map calls into their code, Data-Scientists, ML-Engineers, and Analysts have created programs that handle terabytes of data, and finish running in minutes.

The network filesystem at ./shared makes it trivial to process your data stored in a cloud storage.

from burla import remote_parallel_map

# Run `process_file` on many small machines
results = remote_parallel_map(process_file, files)

# Combine results on one big machine
result = remote_parallel_map(combine_results, [results], func_cpu=64)

The above example demonstrates a basic map-reduce operation.

Burla only takes 2 minutes to try!

Try Burla for free

1. ☝️ Sign in using your Google or Microsoft account.

2. Click the ⏻ Start button to boot some computers.

3. Scale Python over 1,000 CPU's in this Google Colab notebook!

Quick reminder: Burla is open-source and easy to self-host. Click here to deploy Burla in your Cloud.

---

Questions? Schedule a call, or email [email protected]. We're always happy to talk.