Horovod
E760433
Horovod is an open-source distributed deep learning framework designed to make training models across multiple GPUs and machines fast and easy.
All labels observed (1)
| Label | Occurrences |
|---|---|
| Horovod canonical | 1 |
How this entity was disambiguated
This entity first appeared as the object of triple T8823603 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Horovod Context triple: [NCCL, usedBy, Horovod]
-
A.
Farama Foundation
The Farama Foundation is an organization that develops and maintains open-source reinforcement learning tools and libraries for the research and engineering community.
-
B.
NVIDIA Triton Inference Server
NVIDIA Triton Inference Server is an open-source, production-ready platform for serving and scaling AI model inference across GPUs and CPUs with support for multiple frameworks and deployment environments.
-
C.
Pegasos II
Pegasos II is a PowerPC-based computer mainboard developed by Genesi that became popular as a hardware platform for alternative operating systems such as AmigaOS and MorphOS.
-
D.
Toesca
Toesca is an Italian-origin surname best known through Joaquín Toesca, an 18th-century architect influential in colonial Chilean architecture.
-
E.
NVIDIA AI Workflows
NVIDIA AI Workflows are pre-built, end-to-end AI pipelines from NVIDIA that streamline the development, deployment, and scaling of AI applications across common enterprise use cases.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: Horovod Target entity description: Horovod is an open-source distributed deep learning framework designed to make training models across multiple GPUs and machines fast and easy.
-
A.
Farama Foundation
The Farama Foundation is an organization that develops and maintains open-source reinforcement learning tools and libraries for the research and engineering community.
-
B.
NVIDIA Triton Inference Server
NVIDIA Triton Inference Server is an open-source, production-ready platform for serving and scaling AI model inference across GPUs and CPUs with support for multiple frameworks and deployment environments.
-
C.
Pegasos II
Pegasos II is a PowerPC-based computer mainboard developed by Genesi that became popular as a hardware platform for alternative operating systems such as AmigaOS and MorphOS.
-
D.
Toesca
Toesca is an Italian-origin surname best known through Joaquín Toesca, an 18th-century architect influential in colonial Chilean architecture.
-
E.
NVIDIA AI Workflows
NVIDIA AI Workflows are pre-built, end-to-end AI pipelines from NVIDIA that streamline the development, deployment, and scaling of AI applications across common enterprise use cases.
- F. None of above. chosen
Statements (48)
| Predicate | Object |
|---|---|
| instanceOf |
distributed deep learning framework
ⓘ
open-source software ⓘ |
| category |
deep learning software
ⓘ
distributed computing ⓘ machine learning infrastructure ⓘ |
| designGoal |
make distributed deep learning easy to use
ⓘ
make distributed deep learning fast ⓘ |
| developedBy | Uber ⓘ |
| feature |
fault-tolerant training with elastic mode
ⓘ
hierarchical allreduce ⓘ integration with Apache Spark ⓘ mixed-precision training support ⓘ ring-allreduce algorithm for gradient averaging ⓘ timeline profiling for performance debugging ⓘ |
| initialReleaseDate | 2017 ⓘ |
| keyOperation |
allgather
ⓘ
allreduce ⓘ broadcast ⓘ |
| license | Apache License 2.0 ⓘ |
| notableUser | Uber ⓘ |
| optimizedFor |
multi-GPU training
ⓘ
multi-node training ⓘ |
| parallelismType | data parallelism ⓘ |
| primaryUse | distributed training of deep learning models ⓘ |
| programmingLanguage |
C++
ⓘ
CUDA ⓘ
linked to:
NVIDIA CUDA
Python ⓘ |
| repository | https://github.com/horovod/horovod ⓘ |
| supportsFramework |
Apache MXNet
ⓘ
linked to:
MXNet
Keras ⓘ PyTorch ⓘ TensorFlow ⓘ XGBoost ⓘ |
| supportsHardware |
CPU
ⓘ
GPU ⓘ multi-GPU systems ⓘ multi-node clusters ⓘ |
| supportsLanguage |
Apache MXNet
ⓘ
linked to:
MXNet
Keras ⓘ PyTorch ⓘ Spark ML ⓘ
linked to:
Spark MLlib
TensorFlow ⓘ |
| supportsPlatform |
cloud environments
ⓘ
on-premise clusters ⓘ |
| usesCommunicationBackend |
Gloo
ⓘ
MPI ⓘ NCCL ⓘ |
| website | https://horovod.ai ⓘ |
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: Horovod Description of subject: Horovod is an open-source distributed deep learning framework designed to make training models across multiple GPUs and machines fast and easy.
Referenced by (1)
Full triples — surface form annotated when it differs from this entity's canonical label.