Statements (59)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:Library
|
| gptkbp:available_on |
gptkb:Linux
gptkb:mac_OS gptkb:Windows |
| gptkbp:developed_by |
gptkb:Intel_Corporation
|
| gptkbp:enhances |
performance of neural networks
|
| gptkbp:has |
open-source license
|
| https://www.w3.org/2000/01/rdf-schema#label |
Intel DNNL
|
| gptkbp:includes |
fully connected layers
activation functions convolution operations recurrent operations |
| gptkbp:integrates_with |
gptkb:Tensor_Flow
gptkb:MXNet gptkb:Py_Torch |
| gptkbp:is |
cross-platform
scalable widely used designed for efficiency designed for portability designed for flexibility designed for performance community-supported designed for scalability designed for analysis designed for benchmarking designed for deployment designed for evaluation designed for experimentation designed for integration designed for optimization designed for production designed for research designed for simulation designed for testing designed for usability designed for validation designed for visualization highly optimized designed for development designed for profiling designed for verification part of one API |
| gptkbp:is_optimized_for |
gptkb:Intel_architecture
|
| gptkbp:offers |
performance tuning options
|
| gptkbp:provides |
GPU acceleration
CPU acceleration memory management features high-performance primitives |
| gptkbp:released_in |
gptkb:2016
|
| gptkbp:supports |
gptkb:Intel_TBB
gptkb:Intel_AVX-512 gptkb:Intel_MKL deep learning frameworks multi-threading |
| gptkbp:used_in |
machine learning applications
|
| gptkbp:written_in |
gptkb:C++
|
| gptkbp:bfsParent |
gptkb:Intel_DSA
|
| gptkbp:bfsLayer |
6
|