From 89ec05a603e53d555b54b45889696f6a04e182a4 Mon Sep 17 00:00:00 2001 From: Beth Dean Date: Thu, 23 May 2019 15:29:44 -0700 Subject: [PATCH] Update references to VNNI to approved DL Boost nomenclature --- stacks/dlrs/README.md | 2 +- stacks/dlrs/releasenote.md | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/stacks/dlrs/README.md b/stacks/dlrs/README.md index 2e5d2d1..2e8fabf 100644 --- a/stacks/dlrs/README.md +++ b/stacks/dlrs/README.md @@ -1,7 +1,7 @@ # Deep Learning Reference Stack This provides the Deep Learning Reference Stack. To offer more flexibility, there are multiple versions of the Deep Learning Reference Stack: -* Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) primitives and AVX-512 Vector Neural Network Instructions (VNNI) +* Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) primitives and AVX-512 Deep Learning Boost (formerly known as AVX-512 VNNI) * Eigen optimized for Intel Architecture * Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) primitives * OSS PyTorch DLRS Docker image diff --git a/stacks/dlrs/releasenote.md b/stacks/dlrs/releasenote.md index 4f931d9..32c1483 100644 --- a/stacks/dlrs/releasenote.md +++ b/stacks/dlrs/releasenote.md @@ -14,11 +14,11 @@ To offer more flexibility, we are releasing multiple versions of the Deep Learni > **Note:** > For multi-node support, we include a registry with a set of jsonnet files to show integration with Kubeflow for deployment. -## The Deep Learning Reference Stack with Intel® MKL-DNN and VNNI +## The Deep Learning Reference Stack with Intel® MKL-DNN and Intel® AVX512-Deep Learning Boost The release includes: * Clear Linux* OS - * TensorFlow 1.13.1 optimized using Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) primitives and AVX-512 Vector Neural Network Instructions (VNNI) + * TensorFlow 1.13.1 optimized using Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) primitives and Intel® AVX-512 Deep Learning Boost (Formerly Intel® VNNI) * Jupyter Lab* ## The Deep Learning Reference Stack with Eigen @@ -58,7 +58,7 @@ The release includes: ## How to get the Deep Learning Reference Stack -The official Deep Learning Reference Stack Docker images are hosted at: https://hub.docker.com/u/clearlinux/. +The official Deep Learning Reference Stack Docker images are hosted at: https://hub.docker.com/u/clearlinux/. Note that the Intel MKL-DNN-VNNI version is also referred to as the Intel® MKL with AVX-512 Deep Learning Boost in the documentation. * Pull from the [Intel MKL-DNN-VNNI version](https://hub.docker.com/r/clearlinux/stacks-dlrs-mkl-vnni) * Pull from the [Eigen version](https://hub.docker.com/r/clearlinux/stacks-dlrs-oss/) @@ -68,9 +68,9 @@ The official Deep Learning Reference Stack Docker images are hosted at: https:// **Note:** - To take advantage of the AVX-512 and VNNI functionality with the Deep Learning Reference Stack, please use the following hardware: + To take advantage of the AVX-512, and AVX-512 Deep Learning Boost functionality with the Deep Learning Reference Stack, please use the following hardware: * AVX 512 images requires an Intel® Xeon® Scalable Platform - * VNNI requires a Second-Generation Intel® Xeon® Scalable Platform + * AVX-512 Deep Learning Boost requires a Second-Generation Intel® Xeon® Scalable Platform ## Licensing