Dell Technologies ramps up high performance computing, AI prowess
- 20 November, 2019 10:36
Dell Technologies has revealed new solutions, reference architectures and portfolio advancements with integrated technology from partners for improving high performance computing (HPC) and artificial intelligence (AI).
The first of Dell Technologies' announcements, made at the Supercomputing 2019 event in Denver, Colorado, is that the business' suite of Dell EMC Ready Solutions for HPC Storage is getting updated with solutions from partners ThinkParQ’s BeeGFS and ArcaStream’s PixStor file systems.
Dell EMC Ready Solutions for HPC BeeGFS Storage contains ThinkParQ’s software-defined parallel file system, sped-up input/output intensive workloads and the ability to scale from small clusters up to enterprise-class systems located either on premises or in the cloud.
Meanwhile, Dell EMC Ready Solutions for HPC PixStor Storage contains a high-performance parallel file system to allow for data management, archival and analytical functions, a validated and scalable design, object, tape and cloud tiering capabilities through PixStor Ngenea with PowerEdge servers, PowerSwitch and Mellanox networking and PowerVault storage, along with deployment and support services through Dell Technologies.
An Australian use case of these solutions includes the CSIRO's use of the Dell EMC Ready Solutions for HPC BeeGFS Storage for its two petabytes of all NVMe storage.
These Dell EMC Ready Solutions are available now.
In the AI space, Dell EMC has a new validated design for the Domino Data Science Platform, developed in collaboration with partner Domino Data Lab.
The design, tailored towards data scientists, enables users to develop and develop models and also providing information technology (IT) from a centralised platform for the whole data science lifecycle, and will be available globally in December 2019.
Five reference architectures, optimised for Intel Xeon Scalable processors, were also introduced, which includes work from AI-related partners like DataRobot, Grid Dynamics, H20.ai, Iguazio and Kubeflow on Red Hat OpenShift.
These architectures allow for the deployment of AI solutions at a faster rate through Dell EMC converged infrastructure, servers, storage and data protection.
The AI reference architecture for Data Robot, Grid Dynamics, H20.ai and Iguazio will be available sometime this quarter, while Kubeflow on Red Hat OpenShift is available now.
In collaboration with Nvidia, Dell Technologies also announced Nvidia T4 Tensor Core GPUs as an accelerator option for the Dell EMC DSS 8440 server, which contains up to 16 accelerators and uses 70 watts per GPU and is designed for multi-tenancy environments.
Other graphics processing unit (GPU) and field-programmable gate array (FPGA) support for Dell Technologies’ PowerEdge servers include Nvidia Tesla V100S GPU, with up to 25 per cent more bandwidth for communication with Mellanox InfiniBand interconnect and faster data transfers with PowerEdge Express Flash NVMe Performance PCIe SSDs.
The NvidiaTesla V100S GPU and the Intel FPGA Programmable Acceleration Card D5005 are planned to be available through Dell EMC PowerEdge servers in early 2020.
The last of Dell Technologies announcements is the Dell EMC PowerSwitch Z9332F-ON, a 400GbE open networking switch for high performance loads such as cloud service provider data centre networks with intensive compute and storage traffic, like HPC, AI and video streaming.
The Dell EMC PowerSwitch Z9332F-ON is available now.