graphcore-research/tessellate-ipu

TessellateIPU: low level Poplar tile programming from Python

21
/ 100
Experimental

This library helps machine learning researchers and developers, particularly those working with JAX, gain fine-grained control over how their algorithms run on Graphcore IPU hardware. It allows you to specify exactly how data arrays are distributed and processed across the IPU's tiles. You provide your JAX code and data, and it outputs results optimized for IPU's unique architecture, enabling more efficient and customized algorithm execution.

No commits in the last 6 months.

Use this if you are a machine learning researcher or developer using JAX and need to optimize your algorithms by manually controlling data placement and operations on Graphcore IPU hardware tiles.

Not ideal if you are a practitioner looking for a high-level, automated solution for general ML tasks without needing to interact with low-level hardware specifics or if you are not using Graphcore IPUs.

machine-learning-optimization hardware-acceleration scientific-computing high-performance-computing AI-development
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 5 / 25
Maturity 16 / 25
Community 0 / 25

How are scores calculated?

Stars

13

Forks

Language

Python

License

Apache-2.0

Last pushed

Mar 12, 2024

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/graphcore-research/tessellate-ipu"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.