Python* is among the most popular programming languages used for AI, including training machine learning models and performing numerical simulations via determinantal point processes (DPPs)—probability distributions over clouds of points used to model physics, statistics, and machine learning.

The data parallel extension for Python demonstrates high-performing code targeting Intel® hardware—including GPUs and FPGAs—using Python.

The webinar introduces the numba-dpex (Numba Data Parallel Extension) and includes:

• Examples of how to write data parallel code inside @numba.jit decorated and @kernel decorator functions to offload them to a SYCL* device
• Data parallel control (dpctl), which is a companion library intended to make it easier to write extensions that are built into Python and are based on SYCL
• A review of Pairwise distance and K-means use cases to demonstrate the CPU and GPU implementation of numba-dpex

Numba-dpex is packaged as part of Intel® Distribution for Python*.

Skill level: All

Get numba-dpex as part of Intel Distribution for Python, which you can get as a stand-alone version or as part of the AI Tools.