# intel/BigDL

BigDL: Distributed TensorFlow, Keras and PyTorch on Apache Spark/Flink & Ray

Repository: https://github.com/intel/BigDL
Canonical: https://ross.abutalabs.com/products/bigdl
Homepage: https://bigdl.readthedocs.io
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: apache-spark, deep-neural-network, distributed-deep-learning, keras-tensorflow, bigdl, analytics-zoo, python, scala, pytorch
Archived: true
Last push: 2026-06-12T07:14:18+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 87, release rhythm 8, longevity 100
- inputs: {"age_days": 3408, "days_push": 82, "days_rel": 687, "gap_med": null, "n_releases_24m": 1}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2698, forks 730 (observed 2026-08-28T04:07:11.281793+00:00)

## What it is
BigDL is Intel's distributed deep learning library that scales TensorFlow, Keras, and PyTorch workloads on Apache Spark, Flink, and Ray, with sub-libraries for LLM inference, time series (Chronos), recommendations (Friesian), and SGX-secured analytics (PPML). Intel has announced the project will be archived by June 30, 2026, with LLM development moved to IPEX-LLM.

## Use cases
- run distributed pytorch training on apache spark
- scale tensorflow keras pipelines from laptop to cluster
- accelerate deep learning on intel cpus and gpus
- time series forecasting with automl on big data
- build recommendation systems at scale
- run big data ai workloads with sgx hardware security
- distributed llm inference on intel hardware

## When to choose
- you already run Spark/Flink/Ray clusters and need deep learning integrated into them
- you need SGX/TDX-secured big data AI pipelines (PPML)
- you maintain an existing BigDL-based deployment and can fork it

## When to avoid
- starting a new project - Intel is archiving it by 6/30/2026 with no further support
- you only need LLM inference on Intel hardware - use IPEX-LLM instead
- you want actively maintained distributed training - use PyTorch DDP, Ray Train, or Spark-native alternatives

## Facets
- artifact type: library
- maturity: abandoned
- function: deep-learning, machine-learning, llm-training, data-science, etl, streaming
- domain: deep-learning, machine-learning, big-data, data-science, microservices, time-series
- platform: python, jvm, cloud
- tags: apache-spark, ray, distributed-training, tensorflow, pytorch, keras, intel-hardware, analytics-zoo, recommendation-systems, time-series-forecasting, sgx-security, deprecated, gpu, docker

## Member repositories
- intel/BigDL (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:11.281793+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:16:08.512797+00:00, confidence not recorded.
  - readme: https://github.com/intel/BigDL (fetched 2026-08-28T04:07:11.281793+00:00, sha 6f03df8ac7fe)
- Data as of 2026-08-30T08:39:29.467469+00:00.