Slide 36
Slide 36 text
36
Cloud Native Obsidian – Edge AI
The CNO Edge AI named KIRANA is a Computational Storage Solution
that combines CPUs, GPUs, Accelerators with the CNO framework. This
Appliance based solution enables Data Scientists, Machine Learning
Architects of Enterprises and Service Providers to quickly deploy AI/ML
solutions. This solution greatly improves the TCO, Deploy Machine
Learning programs, Deep Learning algorithms, Large Language Models
without the need for complex hardware architectures.
CNO Edge AI
§ Smart Factory, Automation
§ Media & Entertainment
§ Transportation & Logistics
§ Health Care
§ Video Surveillance
USE CASES
CNO framework provides customers to
load their own Kubernetes containers
like TensorFlow, Pytorch, MAPR etc.
GPU enabled frameworks can take full
control of GPUs for their computational
workloads/algorithms
AI/ML
Built-in on Kubernetes, CNO Framework
provides the software infrastructure to
quickly code Machine Learning
programs using JupyterHub and load
frameworks like TensorFlow, Pytorch,
Mathlab etc.
CNO FRAMEWORK
VES Computational Storage Solutions in
a 3U form factor is PCIe Gen5 Hardware
that uses the latest AMD CPUs, NVidia
GPUs, High-Speed Network
Accelerators and High Performance
NVMe Flash SSDs.
HARDWARE
6/20/24 I Viking Enterprise Solutions Confidential