Bringing AI-based Real-time Video Analytics to a new level of scalability, flexibility, and performance
Megh Computing has partnered with Intel and Macnica to create solutions for customers in the areas of Smart Retail, Smart Cities, Smart Buildings, and Smart Factories. These solutions are attractive because Megh has developed the tools, framework, and libraries to generate working systems that can:
- Target a variety of accelerators (CPU, GPU, VPU and FPGA)
- Target a wide range of hardware (Intel NUC to Xeon servers… ask us about IEI platforms)
- Scale readily in performance and streams (and thus a wide range of cost points)
- Cover a broad spectrum of applications
- Map to a variety of platform configurations (edge, private cloud, public cloud and combinations)
- Generate real-time analytics data, then display, report, transmit data and alarm notifications per thresholds as configured
The result is unmatched efficiency, flexibility, performance and cost of ownership.
Built on our best-in-class real-time analytics platform with AI+acceleration, Video Analytics Solution (VAS) is Megh’s real-time streaming analytics platform for generating actionable business intelligence from streaming video for retail, manufacturing, logistics, security, and other segments.
Megh Computing VAS 100 is a complete, end-to-end video real-time analytics platform that offers high performance and low latency. VAS 100 supports up to ten channels of video running on the CPU only. As your analytics needs grow, you can easily upgrade to VAS 300 or higher and accelerate any or all portions of the pipeline using accelerator devices such as GPUs or FPGAs.
Try the VAS-100 Solution for Free!
Get started with scripts that can run on a laptop or PC demonstrating example solutions. You can use video files or tie in IP cameras. We can help you arrange scaled or custom solutions for your application if you choose to take your next step.
VAS 100 is available in two versions for free trail:
VAS 100 AWS
VAS 100 AWS (VAS 100-00 beta) runs on AWS EC2 C5 and T3 instances.
This solution can connect to two remote IP cameras and supports public safety use cases, displaying real-time statistics in a customizable Web dashboard.
Try VAS 100 AWS on AWS Marketplace free for 30 days.
VAS 100 Edge
VAS 100 Edge (VAS 100-10 beta) is deployed on premises on an edge gateway (e.g., an Intel NUC device) in a Docker container.
This solution can connect up to ten cameras and supports public safety use cases.
Try VAS 100 Edge free for 30 days.
Video: VAS-100 Overview
Megh's Nimble Framework & Libraries
Megh’s Nimble™ Framework along with their Arka Libraries and Intel’s OpenVINO™ toolset, allows mapping of pipelines across the compute node (CPU/GPU/VPU/FPGA). The system dynamically assigns inferencing, pre and post processing, and other analytics functions to achieve optimal efficiency. The Megh tools flow includes industry standard libraries (Kafka, OpenCV, gstreamer, etc.) and frameworks.
For those wanting to make customizations, Megh has SDK’s available for a VAS solution or for custom accelerators for FPGA’s. The tools include a shell (Sira Shell) to allow agnostic scale-out services for FPGA accelerators.
Megh’s Deep Learning Engine
Megh Computing has also developed a Deep Learning Engine (DLE) for high-end video analytics systems that can leverage Intel FPGA’s for up to 5x (and soon 10x) GPU performance. The DLE can be integrated in the VAS solution stack or used separately as a custom accelerator. Megh’s DLE SDK may be used to customize accelerators utilizing the DLE.
Contact us today to get started.