amazon-archives/amazon-dsstne

Health v2 · maintenance only

Flags: no_releases archived

Adoption not part of the score

4390 stars · 723 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

DSSTNE (Deep Scalable Sparse Tensor Network Engine) is a C++ GPU library from Amazon for training and deploying deep learning recommendation models with sparse inputs and outputs. It uses model-parallel multi-GPU scaling to handle weight matrices too large for a single GPU.

Use cases

• train recommendation models on sparse data
• scale deep learning training across multiple GPUs
• build personalized product recommendation engines
• train neural networks with layers too large for one GPU
• generate recommendations at production scale

When to choose

• you need fast training on very sparse recommendation datasets
• your model's weight matrices exceed single-GPU memory
• you want production-oriented speed over experimental flexibility

When to avoid

• you need actively maintained software with recent updates
• you want flexible experimentation with modern architectures
• you need general-purpose deep learning beyond recommendation workloads
• you prefer a rich ecosystem like PyTorch or TensorFlow

Facets

library · maturity abandoned

machine-learning deep-learning gpu-computing machine-learning deep-learning large-language-models cpp recommendation-systems sparse-tensors model-parallelism multi-gpu cuda linux gpu docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="amazon-archives/amazon-dsstne")