Enhancing the Dell Validated Design for Computer Vision with BOSS AI on Tanzu Kubernetes Grid
The first two releases of the Dell Validated Design for Computer Vision platform focused on integrating a single VMS product with multiple computer vision applications developed by different ISVs. All applications were implemented using server virtualization with VMware for 100% of the required components. We tested selected functionality, performance, and high availability management. A larger percentage of the testing time was devoted to analyzing the scalability of the NVIDIA A40 GPU for workloads that require video encoding or decoding plus machine learning object detection model inference on every video frame. The third release of the platform design added hosting security threat detection and remediation using on-premises application components from in a multicloud, distributed architecture connected to managed services running on the AWS hyperscale platform. We tested all threat detection and remediation aspects, including scanning and updating nontraditional IT assets like IP cameras used in video monitoring.