PlaidML
E99362
PlaidML is an open-source, hardware-agnostic deep learning engine designed to accelerate neural network computation on a wide range of GPUs and other devices.
Statements (49)
| Predicate | Object |
|---|---|
| instanceOf |
deep learning engine
→
machine learning library → open-source software → tensor compiler → |
| acquiredBy |
Intel
→
|
| canBeIntegratedWith |
existing deep learning workflows
→
|
| category |
AI framework
→
GPU computing library → |
| designedFor |
deep learning acceleration
→
neural network computation → |
| developedBy |
Vertex.AI
→
|
| feature |
automatic kernel generation
→
backend-agnostic computation → cross-platform support → graph optimization → tensor algebra compiler → |
| goal |
avoid vendor lock-in
→
provide high-performance deep learning on diverse hardware → |
| hasComponent |
Tile IR
→
backend code generators → |
| implements |
tile language
→
|
| isHardwareAgnostic |
true
→
|
| maintainer |
Intel
→
|
| optimizationTarget |
performance
→
portability → |
| programmingLanguage |
C++
→
Python → |
| repositoryPlatform |
GitHub
→
|
| softwareLicense |
Apache License 2.0
NERFINISHED
→
|
| supportsBackend |
LLVM
→
Metal → OpenCL → |
| supportsFramework |
Keras
NERFINISHED
→
ONNX → |
| supportsHardwareType |
AMD GPU
→
CPU → GPU → Intel GPU → Metal devices → NVIDIA GPU NERFINISHED → OpenCL devices → |
| supportsLanguage |
C++
→
Python → |
| supportsOS |
Linux
→
Windows → macOS → |
| useCase |
inference of neural networks
→
running Keras models on non-CUDA GPUs → training neural networks → |
Referenced by (1)
| Subject (surface form when different) | Predicate |
|---|---|
|
Keras
→
|
supportsBackend |