Back to Table of contents

Primeur live 2018-11-14

Quantum computing

Atos delivers world's highest performing quantum simulator to Argonne National Laboratory  ...

Focus on Europe

Barcelona Supercomputing Center releases COMPSs version 2.4 at SC18 ...

Middleware

Spectra7 technology now available from 10Gtek  ...

ScaleMP to showcase broad product line-up and innovations at SC18 ...

OpenMP 5.0 is a major leap forward ...

Excelero announces support for Mellanox's BlueField, taking web-scale storage with NVMesh to a whole new level ...

Mellanox HDR InfiniBand solutions accelerate new supercomputer for advanced research computing at the University of Michigan ...

Hardware

Ciphertex debuts two new servers at Supercomputing'18 ...

TYAN highlights AMD EPYC processor-based HPC and storage server platforms at SC18 ...

Liqid delivers NVIDIA GPU supercomputer rack as it ramps innovation in bare-metal composable infrastructure ...

Liqid enables industry's first composable FPGA platform as it broadens composable infrastructure leadership ...

Applications

Mellanox and Los Alamos National Laboratory join forces to develop ultra-large scale mission-centric computing infrastructure ...

TOP500

Mellanox InfiniBand and Ethernet solutions accelerate the majority of TOP500 platforms on the November TOP500 supercomputers list ...

The Cloud

Cray partners with Microsoft and Intel to deliver early-access programme for running mission-critical supercomputing workloads in Microsoft Azure ...

Univa accelerates enterprise migration of HPC workloads to the Cloud ...

WekaIO and Univa partner to simplify the on-ramp of HPC workloads to the Cloud ...

Company news

Atos launches new BullSequana hybrid supercomputer for AI-augmented simulation ...

Asetek Liquid Cooled Intel Compute Modules installed by Ace Computers in HPC cluster for U.S. Defense contractor ...

Asetek to showcase liquid cooling technology for AI and HPC at SC18 ...

Verne Global introduces Arm architecture via Marvell ThunderX2 processor-based servers to hpcDIRECT ...

Cray turbocharges AI model development with new algorithms and frameworks ...

Solarflare, Smart IOPS and Wistron showcase high performance NVMe over TCP storage appliance at Supercomputing 2018 ...

Latest version of Bright Cluster Manager deployed at Binghamton University by Data In Science Technologies ...

DDN unveils professional support for Lustre clients on Arm-based platforms ...

Panasas introduces ActiveStor Ultra featuring PANFS 8, the industry's only plug-and-play portable parallel file system ...

Cray turbocharges AI model development with new algorithms and frameworks

12 Nov 2018 Seattle - Cray Inc. has added enhancements to its Urika AI and Analytics software suites, adding tools that enable data scientists to train artificial intelligence (AI) models more accurately and in less time. New features in the Cray Urika-CS and Urika-XC AI and Analytics suites include Cray-developed libraries to intelligently optimize machine learning model settings as well as additional AI tools and frameworks commonly used by data scientists.

One of the most challenging tasks for a data scientist is optimizing their choice of model hyperparameters, the knobs they can tune to pick the best model within a model class. This optimization can be resource- and labour-intensive, and often relies on time-consuming hand-crafted or brute-force approaches. Cray is adding hyperparameter optimization (HPO) algorithms, capable of running in a distributed fashion, to help data scientists find the optimal model for production AI deployments.

"Developing AI models can be a complex, time-consuming process. By offering intelligent hyperparameter optimization support within our Urika-CS and Urika-XC suites, we're giving data scientists pre-set algorithms to quickly identify the most favorable machine learning model designs", stated Per Nyberg, vice president market development, AI and cloud at Cray. "Offering data scientists this support has several benefits, including increased productivity and workflow efficiencies."

The new Urika suites are augmented with four HPO strategies - two commonly used strategies and two strategies developed by Cray to take advantage of the parallelism available in a distributed system. Taken together, these strategies simplify the task of finding and tuning the optimal model for a given application.

The four strategies are:

  • Genetics-based: to find more optimal model architectures
  • Population-based: to find the best way to train your model faster
  • Random: a baseline algorithm to guide behaviour
  • Grid Search: a traditional approach, guided by performance metrics

Cray has added four popular analytics and deep learning frameworks to its Urika AI and Analytics suites: PyTorch, Keras and Horovod for model development and training, as well as the highly-scalable Programming Big Data in R (pbdR) package. The upgraded software suites provide researchers and data science teams the right tools and more choice for how they perform their analytics, machine learning and deep learning workflows.

The new versions of the Urika-CS AI and Analytics software suite and Urika-XC software suite are expected to be available within 90 days.

To learn more and to see a live demo of these new capabilities, you can stop by the Cray booth #2413 at SC18.

Source: Cray

Back to Table of contents

Primeur live 2018-11-14

Quantum computing

Atos delivers world's highest performing quantum simulator to Argonne National Laboratory  ...

Focus on Europe

Barcelona Supercomputing Center releases COMPSs version 2.4 at SC18 ...

Middleware

Spectra7 technology now available from 10Gtek  ...

ScaleMP to showcase broad product line-up and innovations at SC18 ...

OpenMP 5.0 is a major leap forward ...

Excelero announces support for Mellanox's BlueField, taking web-scale storage with NVMesh to a whole new level ...

Mellanox HDR InfiniBand solutions accelerate new supercomputer for advanced research computing at the University of Michigan ...

Hardware

Ciphertex debuts two new servers at Supercomputing'18 ...

TYAN highlights AMD EPYC processor-based HPC and storage server platforms at SC18 ...

Liqid delivers NVIDIA GPU supercomputer rack as it ramps innovation in bare-metal composable infrastructure ...

Liqid enables industry's first composable FPGA platform as it broadens composable infrastructure leadership ...

Applications

Mellanox and Los Alamos National Laboratory join forces to develop ultra-large scale mission-centric computing infrastructure ...

TOP500

Mellanox InfiniBand and Ethernet solutions accelerate the majority of TOP500 platforms on the November TOP500 supercomputers list ...

The Cloud

Cray partners with Microsoft and Intel to deliver early-access programme for running mission-critical supercomputing workloads in Microsoft Azure ...

Univa accelerates enterprise migration of HPC workloads to the Cloud ...

WekaIO and Univa partner to simplify the on-ramp of HPC workloads to the Cloud ...

Company news

Atos launches new BullSequana hybrid supercomputer for AI-augmented simulation ...

Asetek Liquid Cooled Intel Compute Modules installed by Ace Computers in HPC cluster for U.S. Defense contractor ...

Asetek to showcase liquid cooling technology for AI and HPC at SC18 ...

Verne Global introduces Arm architecture via Marvell ThunderX2 processor-based servers to hpcDIRECT ...

Cray turbocharges AI model development with new algorithms and frameworks ...

Solarflare, Smart IOPS and Wistron showcase high performance NVMe over TCP storage appliance at Supercomputing 2018 ...

Latest version of Bright Cluster Manager deployed at Binghamton University by Data In Science Technologies ...

DDN unveils professional support for Lustre clients on Arm-based platforms ...

Panasas introduces ActiveStor Ultra featuring PANFS 8, the industry's only plug-and-play portable parallel file system ...