Archived AWS Neuron SDK documentation#
Note
This page contains archived tutorials and other documentation for older versions of the AWS Neuron SDK. These pages are no longer actively maintained and may reference unsupported features or deprecated APIs. They are provided as-is and may not reflect the current state of the AWS Neuron SDK.
Overview#
The following content has been archived for reference purposes. For the latest documentation and guides, visit the AWS Neuron SDK documentation.
Archived feature docs#
Feature |
Last release supported |
Date archived |
|---|---|---|
Neuron 2.28.0 |
Archived on: 5/11/2026 |
|
Neuron 2.28.0 (Trn1 plugin); Neuron 2.27.0 (Inf1 plugin) |
Archived on: 5/11/2026 |
|
Neuron 2.27.0 |
Archived on: 12/2/2025 |
|
Neuron 2.27.0 |
Archived on: 12/2/2025 |
|
Neuron 2.25.0 |
Archived on: 9/15/2025 |
|
Neuron 2.22.0 |
Archived on: 7/24/2026 |
|
Neuron 2.27.0 |
Archived on: 3/30/2026 |
|
Neuron 2.16.0 |
Archived on: 3/11/2026 |
|
Neuron 2.22.0 |
Archived on: 3/11/2026 |
|
Neuron 2.22.0 |
Archived on: 3/11/2026 |
|
Mixed — legacy per-component app notes |
Archived on: 7/21/2026 |
|
Neuron 2.32.0 |
Archived on: 7/24/2026 |
Archived deployment flows#
The following deployment flow pages were archived during the consolidation of /containers/, /devflows/, and /dlami/ into /deploy/. These pages reference legacy patterns (Inf1, Neo compilation, BYOC hosting) that have been superseded by current deployment guides.
Page |
Reason archived |
Date archived |
|---|---|---|
Bring Your Own Neuron Container to SageMaker Hosting (Inf1) — Archived |
Legacy BYOC hosting flow |
Archived on: 4/20/2026 |
Bring Your Own Neuron Container to SageMaker Hosting (Inf2 or Trn1) — Archived |
Legacy BYOC hosting flow |
Archived on: 4/20/2026 |
Legacy SageMaker hosting flow |
Archived on: 4/20/2026 |
|
Compile with Sagemaker Neo and Deploy on Sagemaker Hosting (inf1) |
Legacy Neo compilation flow |
Archived on: 4/20/2026 |
Superseded by EKS-specific guides |
Archived on: 4/20/2026 |
|
Legacy runtime 1.0 |
Archived on: 4/20/2026 |
Archived quickstarts and getting-started pages#
The following pages were archived when the quickstarts moved to the TorchNeuron, vLLM Neuron, vLLM Omni Neuron, NKI, Neuron Explorer, and Neuron Observer stack in Neuron 2.33.0. For current quickstarts, see Get started with AWS Neuron.
Page |
Reason archived |
Date archived |
|---|---|---|
Uses Trn1. TorchNeuron supports only Trn2 and Trn3. |
Archived on: 10/13/2026 |
|
Uses the legacy PyTorch/XLA |
Archived on: 10/13/2026 |
|
Superseded by Neuron agentic development |
Archived on: 10/13/2026 |
Archived samples, benchmarks, FAQs, and tools from About Neuron#
The following pages were moved from the About Neuron section because they describe retired instances, frameworks, or flows.
Page |
Reason archived |
Date archived |
|---|---|---|
Lists samples for transformers-neuronx, NxD Training, NeMo, torch-neuron, TensorFlow Neuron, and MXNet Neuron |
Archived on: 10/13/2026 |
|
Inf1, Inf2, and Trn1 benchmarks measured with retired or legacy components (TensorFlow Neuron, torch-neuron, transformers-neuronx, PyTorch/XLA, NxD Core, and NxD Training) |
Archived on: 10/13/2026 |
|
ONNX (torch-neuron), Inf1 inference, Inf1 troubleshooting, and Trn1 training FAQs |
Archived on: 10/13/2026 |
|
transformers-neuronx tensor parallel calculator for Inf2 and Trn1 |
Archived on: 10/13/2026 |
|
Applies only to Inf2 and Trn1; NKI is the custom kernel path on Trn2 and Trn3 |
Archived on: 10/13/2026 |
|
|
Archived on: 10/18/2026 |
Legacy setup#
Setup and installation guides for unsupported install targets and superseded Neuron releases have been archived. See the Archived setup guides landing page for the full list — including Rocky Linux 9, Ubuntu 20.04, and Amazon Linux 2 install targets, and previous-release PyTorch NeuronX installs (Neuron 2.7.0–2.9.0).
Archived tutorials#
Tutorial |
Last release supported |
Date archived |
|---|---|---|
Neuron 2.24.0 |
Archived on: 7/31/2025 |
|
Neuron 2.24.0 |
Archived on: 7/31/2025 |
|
Neuron 2.25.0 |
Archived on: 7/31/2025 |
|
Fine-tuning Llama2 7B with tensor parallelism and ZeRO-1 optimizer using Neuron PyTorch-Lightning |
Neuron 2.26.0 |
Archived on: 8/25/2025 |
Neuron 2.26.0 |
Archived on: 8/25/2025 |
|
Training CodeGen2.5 7B with Tensor Parallelism and ZeRO-1 Optimizer |
Neuron 2.26.0 |
Archived on: 8/25/2025 |
Neuron 2.26.0 |
Archived on: 8/25/2025 |
|
Neuron 2.29.0 |
Archived on: 3/30/2026 |
|
Neuron 2.28.0 |
Archived on: 5/11/2026 |
|
Track Training Progress in TensorBoard using PyTorch Neuron (archived) |
Neuron 2.28.0 |
Archived on: 5/11/2026 |
Accessing Archived Content#
Each tutorial listed above corresponds to a specific version or feature set of the Neuron SDK that has since been superseded. Use these resources for historical context or migration guidance.
Warning
Archived tutorials may not be compatible with current Neuron SDK releases. Exercise caution when following instructions from these documents.