🚀 Infrastructure

Our HPC Lab is built with high-performance hardware to support parallel computing and large-scale research:

• 🔹 40 Slave Nodes working in parallel

• 🔹 2 Master Nodes managing and coordinating tasks

• 🔹 256 GB RAM per node for high-speed data processing

• 🔹 Intel Xeon Processors for powerful computation

• 🔹 NVIDIA K40m Graphics Card for accelerated computing

💡 The master node divides large problems into smaller tasks, distributes them to slave nodes, and combines results efficiently to deliver fast and accurate outputs. 

⚙️ Detailed Configuration

🧠 Master Node

• CPU: Intel Xeon E5-2670 v3 @ 2.30GHz  (Total Cores: 24 (2 Sockets × 12 Cores) )

• Dual Intel Xeon E5-2670 v3 Processors

• 256 GB RAM

• RAID Controller, Dual Port FDR Adapter, Dual Port GbE

• GPU: No GPU (CPU-only node) 

• Storage:

  1. Root (/): 772 GB 

  2. Shared /home (NFS): 6.8 TB 

Software Configuration

• Operating System: Rocky Linux 8.8 (Green Obsidian)

• Kernel: EL8 compatible

• Workload Manager: Slurm Controller (slurmctld)

💡 Role Insight:

The master node acts like the conductor of an orchestra, assigning tasks, managing users, and coordinating execution across all compute nodes. 

⚙️ Slave Node

• IBM nx360 M5 Server

• CPU: Intel Xeon E5-2670 v3 @ 2.30GHz  (Total Cores: 24 (2 Sockets × 12 Cores) )

• 256 GB RAM

• 2 × 900GB SAS 10K HDDs

• RAID Controller

• Dual Port FDR Adapter

• Dual Port GbE

• 🎮 GPU: 2 × NVIDIA Tesla K40m (12 GB GDDR5 Memory each) per node (High-performance GPU acceleration) 

• Storage: Shared /home via NFS (6.8 TB total capacity)

Software Configuration

• Operating System: Rocky Linux 8.8

• Kernel Version: 4.18.0-477.27.1.el8_8.x86_64

• Workload Manager: Slurm Worker Daemon (slurmd)

⚡ Processing Insight:
These nodes are the workhorses of the HPC cluster, executing jobs in parallel. The dual Tesla K40m GPUs per node significantly accelerate workloads like:

• Deep Learning

• Scientific Simulations

• Large-scale Data Processing

💾 Storage System (Shared Across Cluster) 

• IBM x3650 M4 Rack Server

• Dual Intel Xeon E5-2609 v2 Processors

• 64 GB ECC DDR3 RAM

• Storage Configuration:

  • 2 × 600GB SAS 10K RPM HDDs

  • 6 × 1.8TB SAS HDDs

• RAID Controller

• Dual Port FDR Adapter

• 2 × 8Gb FC HBAs

🚀 Overall Resources – HPC Cluster

🧠 Master Nodes (2 Total)

• 🔹 Total CPU Cores: 48 Cores (24 × 2 nodes)

• 🔹 Total RAM: 512 GB (256 GB × 2 nodes)

💡 Role: Centralized scheduling, job management, and orchestration using Slurm Controller (slurmctld)

⚙️ Compute Nodes (40 Total)

• 🔹 Total CPU Cores: 960 Cores (24 cores × 40 nodes)

• 🔹 Total RAM: 10.24 TB (256 GB × 40 nodes)

🎮 GPU Power (NVIDIA Accelerated Computing)

• 🔹 GPU Model: NVIDIA Tesla K40m

• 🔹 GPUs per Node: 2

• 🔹 Total GPUs: 80 GPUs (2 × 40 nodes)

🔥 GPU Memory Power

• 🔹 Per GPU: 12 GB GDDR5

• 🔹 Per Node: 24 GB GPU Memory

• 🔹 Total GPU Memory: 960 GB (24 GB × 40 nodes)

🌟 In One Line

👉 A powerful fusion of 960 CPU cores + 80 GPUs + 10.24 TB RAM — built to turn complex ideas into real-world breakthroughs.

🔬 Research Domains

The HPC Lab supports a wide range of research areas, including:

• 🧪 Science & Engineering Simulations

• 🌾 Agriculture & Crop Analysis

• 💊 Medical Research & Data Processing

• 🤖 Artificial Intelligence & Machine Learning

• 📊 Big Data Analytics

This enables researchers to experiment, simulate, and innovate at scale.

🎯 Objectives

• Enable high-speed parallel computing for research

• Support large-scale data analysis and simulations

• Provide infrastructure for AI and GPU-based computing

• Foster innovation and interdisciplinary collaboration

🌍 Impact

This HPC Lab acts as a research accelerator, empowering students, faculty, and researchers to solve real-world problems efficiently.

From improving agricultural productivity to advancing healthcare analytics, the lab transforms ideas into impactful solutions.

📞 Get Involved

Interested in using the HPC Lab for research or collaboration?

📧 Fill up google form by using provided below link
🏫 Visit: Ganpat University Campus