Rocket CE for Power

Below are the FAQ's which will provide the quick info on RocketCE How RocketCE conda packages are useful? Rocket provides conda packages for Machine Learning through RocketCE. These conda packages are built and provided for installing on Power 9 and Power 10 machines having ppc64le architecture. Where are RocketCE conda packages are located? RocketCE conda packages are available in anaconda rocketce channel (https://anaconda.org/rocketce/repo) . What versions of Python are supported for RocketCE conda packages? Presently RocketCE conda packages are built on OpenCE version of 1.3.1 and is available and supported for Python 3.8. What conda packages are available for RocketCE? RocketCE provides the following packages under anaconda rocketce channel Tensorflow Pytorch XGBoost OnnxRuntime Are the RocketCE conda packages are available for both CPU and GPU? Present release of RocketCE for OpenCE 1.3.1 conda packages are available which supports only CPU. GPU supported packages wil

The matrix-multiply assist (MMA) facility was introduced by the Power Instruction Set Architecture (ISA) v3.1. The related instructions implement numerical linear algebra operations on small matrices and are meant to accelerate computation-intensive kernels, such as matrix multiplication, convolution and discrete Fourier transform. To efficiently accelerate MMA operations, the Power10 processor core implements a dense math engine (DME) microarchitecture which effectively provides an accelerator for cognitive computing, machine learning and artificial intelligence inferencing workloads. The DME encapsulates compute efficient pipelines, a physical register file, and associated data-flow that keeps resulting accumulator data local to the compute units. Each MMA pipeline performs outer-product matrix operations, reading from and writing back a 512-bit accumulator register. Power10 implements the MMA accumulator architecture without adding an additional architected state. Each architected 5

WML CE in OpenShift Follow these steps to deploy IBM Watson® Machine Learning, Distributed Deep Learning (DDL) directly into your enterprise private cloud with Red Hat OpenShift 3.x by using TCP or InfiniBand communication between the worker nodes. Before you begin Install the OpenShift and Helm CLI to deploy your application from the command line. After installing the CLI, add the IBM® Helm Chart repository. Refer to the wmlce-openshift README for more details Deploying WML CE DDL with TCP cross node communication Create container SSH keys as a Kubernetes secret. mkdir -p .tmp yes | ssh-keygen -N "" -f .tmp/id_rsa oc create secret generic sshkeys-secret --from-file=id_rsa=.tmp/id_rsa --from-file=id_rsa.pub=.tmp/id_rsa.pub Deploy the WML CE OpenShift Helm Chart with DDL enabled: helm install --name ddl-instance --set license=accept Path_of_Chart --set resources.gpu=8 --set paiDistributed.mode=true --set paiDistributed.sshKeySecret=sshkeys-secret name The deployment name

Rocket is providing a conda channel for downloading, installing, and upgrading all Power open source Cognitive Environment components supported by Rocket.  With conda, a single command can replace dozens of steps and commands previously required to install some open source products.  For example, installing TensorFlow with its dependencies formerly required over thirty steps but now can be accomplished with a single conda command. Conda is an open source tool ideally suited for delivery and deployment of software due to its channel concept and multiple types of channels supporting varying degrees of security including file channels for air-gapped systems. Rocket has set up two channels from which you can download OpenCE open source ports using conda – one public and open to all, the other a secure channel available only to Rocket customers on a support plan. To install Anaconda or Miniconda on your Power-based Linux see:  https://conda.io/projects/conda/en/latest/user-guide/install/ind