NVIDIA DGX Supercomputing Cluster
NVIDIA DGX systems, including the university's dedicated DGX Spark units, provide high-throughput acceleration for faculty research, Ph.D. dissertations, and student projects. The cluster handles massive parallelized backpropagation for vision transformers, diffusion models, and foundation LLMs.
AI & Robotics Lab (VV126)
Home of SARADHI AI and physical edge experimentation in Vidyut Vihar. The facility hosts the NVIDIA Jetson edge server fleet powering SARADHI AI SchoolBox. Development benchmarking turned a 10-minute lesson audio recording into 25 contextual quiz items in ~1.5 minutes.
AI & ML Distributed Edge-Cloud Hybrid Topology
How heavy training on DGX clusters connects seamlessly with lightweight offline inferencing at the classroom edge.
NVIDIA DGX Spark units compute multi-billion parameter foundation checkpoints.
Cloud ClusterKubernetes container deployment at saradhi.co.in serving web users.
Cloud ServiceQuantized ONNX models running offline inside rural classrooms with zero latency.
Local OfflineStudent mobile quizzes, audio assessments, and DICOM diagnostic second opinions.
Realtime Impact