Pre-Release and NDA Containers

Published: 04/13/2021

Intel is committed to providing world class containers, models, and resources through the oneContainer Portal.  Here you’ll find containers to help drive innovation and see your Intel® XPU based project into production. The containers included in this download require a Non-Disclosure Agreement and an appropriate Intel account to download, as they may include Intel-confidential and pre-release content.  


Intel XE HP
Xe-HP SDV Containers NDA  
Request Access 

Already have access? 
Sign In 


The following toolkits are available for NDA use with the Xe-HP SDV:

  • Intel® oneAPI Base Toolkit which contain many core libraries and optimization tools
  • Intel®, the Intel® oneAPI HPC Toolkit containing the key compilers for HPC applications
  • Intel® oneAPI AI Analytics Toolkit containing versions of TensorFlow* and PyTorch* frameworks

Learn more about Intel® oneAPI Toolkits (including downloading the public toolkit versions for production XPUs)

Browse All Containers

 

AI Solutions NDA Container Build Archives 

Compute-intensive AI workloads perform inference and training tasks that often scale beyond a single processor to multiple machines. Intel's AI container solutions simplify this scaling for cloud and on-premises deployments.  These NDA solutions utilize TensorFlow* and PyTorch* frameworks and offer models for classification and language processing workloads 

Topology 

Framework 

Precision 

ResNet50v1.5 

TensorFlow 

BF16 

ResNet50v1.5 

TensorFlow 

BF16 

BERT-LG 

TensorFlow 

BF16 

BERT-LG 

TensorFlow 

BF16 

BERT-LG 

TensorFlow 

FP16 

BERT-LG 

PyTorch 

FP16 

ResNet50v1.5 

PyTorch 

BF16 

ResNet50v1.5 

PyTorch 

BF16 

ResNet50v1.5 

TensorFlow 

Int8 

ResNet50v1.5 

PyTorch 

Int8 

DLRM 

PyTorch 

FP16 

 

Product and Performance Information

1

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.