Dell, VMware, and NVIDIA are committed to delivering IT platforms that are both highly performant and easy to manage. We have delivered a series of validated design guide releases based on engineering lab projects using real-world applications, including VMS, CV, enterprise systems security, and data science. There is currently ample evidence proving that it is not only possible but preferable to host multiple large-scale virtualized applications on a single VxRail cluster equipped with NVIDIA GPUs and managed with VMware TKG and vSphere.
Our latest testing showed that we could start with the same hardware and management tools used in our first release described previously and extend the capability to support containerized applications using TKG. VMware TKG was configured to run on an existing instance of our original Dell Validated Design for Computer Vision platform infrastructure to include support for containerized applications. The process for adding this functionality is well-documented and fully supported by all three key solution partners.
To test the capability of this new TKG environment on our updated VxRail cluster, we implemented the BOSS AI Enterprise AI platform. BOSS provides world-class advanced predictive analytic capabilities in a Low-Code/No-Code environment while supporting both structured and unstructured data. The BOSS platform empowers companies to build secure inter/intra-company predictive models without transporting secret or private data across the network. Our validation demonstrates how to add data science data management and model-building capability to the rich set of virtualized VMS and CV applications to support additional workflow enhancements.
Our final validation test involved building an enhanced operational workflow that combined insights from a CV application with simulated data representing badge use data that would typically be available from a building management system. We used BOSS data ingestion for the two datasets and then joined the information into a single BOSS Virtual Data Set. Data science professionals frequently spend too much of their effort performing these data engineering and data management tasks on every project. The convenience of having a standard framework running on TKG-managed containers will improve the productivity of any data science team regardless of their level of maturity.
Organizations that are modernizing their hardware platforms and software systems for large-scale operations management need many different types of applications. Some will come deployable as virtual machines, and some will use container virtualization. Dell Technologies, VMware, and NVIDIA have invested in designing and testing a single common platform capable of providing the compute, GPU acceleration, storage performance, and flexibility to host most or all the needs of even the largest operational environments.