liveBook · Manning

Parallel map

A parallel map is an operation that applies a computation concurrently across inputs or processing devices and returns the corresponding results.

How 2 Manning books explain it

Deep Learning with JAX

The JAX mapping transformation that parallelizes a computation across multiple local devices.

How it is used
Parallelizes independent chunks across local accelerators for SPMD computation, including data-parallel neural-network training.
How deep it goes
Explains device setup, input reshaping, mapping parameters, collectives, nested transformations, training, and multihost execution.
Examples
Demonstrates parallel dot products, CPU device emulation, reshaped vector batches, and data-parallel MNIST-style training.
What you can do after
Apply pmap() to split array workloads across local devices and aggregate results or gradients with collective operations.

Read how Deep Learning with JAX explains parallel map →

Clojure: The Essential Reference

pmap is a Clojure function that applies a function to one or more collections in parallel and returns a lazy sequence of results in input order.

How it is used
Applies independent collection transformations in parallel while retaining lazy, ordered sequence consumption.
How deep it goes
Explains contracts, laziness, ordering, chunking, exceptions, and performance tradeoffs, but does not provide a transducer form.
Examples
Uses elementwise addition, delayed printing, uneven sleeps, chunked evaluation, and Criterium benchmarks.
What you can do after
Apply pmap to suitable independent workloads and evaluate whether chunking or sequential map better fits the cost profile.

Read how Clojure: The Essential Reference explains parallel map →

Related concepts

These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.