Products - VISARJ DCBx Immersion Cooling Solutions

Complete Range of Immersion-Cooled Pods

Each VISARJ DCBx pod is engineered for specific deployment scenarios, offering unmatched efficiency, density, and operational simplicity. Choose the configuration that matches your capacity, deployment timeline, and use case requirements.

Quick Comparison

Specification	12U Pod	24U Pod	42U Pod
Form Factor	900mm × 1400mm × 1100mm	720mm × 1500mm × 1100mm	720mm × 2500mm × 1100mm
Capacity	15-40 kW	50-100 kW	100-200 kW
Deployment Time	<8 weeks	8-10 weeks	10-12 weeks
Rack Configuration	Up to 12 dual-socket servers	Up to 24 dual-socket servers	Up to 42 dual-socket servers
Fluid Capacity	~220-250 Liters	~450-500 Liters	~950-1,000 Liters
Best For	Edge, Remote Sites	Regional Centers, R&D	Core Infra, AI/ML

Compact Edge Solution

12U Pod – Compact Edge Pod

Technical Specifications

Capacity: 15-40 kW
Form Factor: 900mm(W) × 1400mm(D) × 1100mm(H)
Deployment: <8 weeks
Dielectric Fluid Capacity: ~220-250 Liters
Use Case: Ultra-compact, ideal for distributed edge or on-prem sites
Pod weight (empty): ~150kgs
Pod weight (with fluid and servers): ~450-500kgs
Rack configuration: Up to 12 dual-socket servers (1U)

Detailed Configuration

Server Configuration

Quantity: Up to 12 servers
Form Factor: 1U dual-socket
CPU Options: Intel Xeon Scalable or AMD EPYC
Memory: DDR4/DDR5, up to 2TB per server
Storage: NVMe SSD, hot-swappable

Networking

Switch Type: 1U managed switch
Port Options: 1/10/25/100 GbE
Redundancy: Dual uplinks
Management: Out-of-band access

Power & Cooling

Power Input: 3-phase 415V AC
Redundancy: N+1 PSU configuration
Cooling Method: Single-phase immersion
Fluid Type: Dielectric mineral oil
PUE: < 1.1 typical

Monitoring & Management

BMC: IPMI 2.0 / Redfish API
Sensors: Temperature, fluid level, power
Alerts: SNMP, syslog, email
Dashboard: Web-based management console

Use Cases

Edge AI processing, telecom rollouts, remote research labs, branch or campus IT.

Value Proposition

Lightweight, fits into existing electrical grids, 3-phase ready.

Mid-Range Deployment

24U Pod – Mid-Range Deployment

Technical Specifications

Capacity: 50-100 kW
Form Factor: 720mm(W) × 1500mm(D) × 1100mm(H)
Deployment: 8-10 weeks
Dielectric Fluid Capacity: ~450-500 Liters
Use Case: Balanced footprint for enterprises or regional clusters
Pod weight (empty): ~300kgs
Pod weight (with fluid and servers): ~1,100-1,200kgs
Rack configuration: Up to 24 dual-socket servers (1U)

Detailed Configuration

Server Configuration

Quantity: Up to 24 servers
Form Factor: 1U dual-socket
CPU Options: Intel Xeon Scalable or AMD EPYC
Memory: DDR4/DDR5, up to 2TB per server
Storage: NVMe SSD, hot-swappable
GPU Support: Optional accelerator cards

Networking

Switch Type: 1U ToR managed switch
Port Options: 10/25/100 GbE
Redundancy: Dual redundant switches
Uplinks: 100 GbE aggregated
Management: Out-of-band + VLAN support

Power & Cooling

Power Input: 3-phase 415V AC
Distribution: Redundant PDUs
Redundancy: N+1 PSU configuration
Cooling Method: Single-phase immersion
Fluid Type: Dielectric mineral oil
PUE: < 1.1 typical

Monitoring & Management

BMC: IPMI 2.0 / Redfish API
Sensors: Temperature, fluid level, power, flow rate
Alerts: SNMP, syslog, email, SMS
Dashboard: Advanced web-based console
Integration: DCIM/monitoring tools support

Use Cases

Regional financial centers, R&D workloads, manufacturing sites, disaster recovery pods.

Value Proposition

Flexible deployment at edge or core with balanced density and scalability.

Full Rack Solution

42U Pod – Full Rack Pod

Technical Specifications

Capacity: 100-200 kW
Form Factor: 720mm(W) × 2500mm(D) × 1100mm(H)
Deployment: 10-12 weeks
Dielectric Fluid Capacity: ~950-1,000 Liters
Use Case: Standard high-density pod for core or large-scale deployments
Pod weight (empty): ~600kgs
Pod weight (with fluid and servers): ~2,100-2,200kgs
Rack configuration: Up to 42 dual-socket servers (1U)

Detailed Configuration

Server Configuration

Quantity: Up to 42 servers
Form Factor: 1U dual-socket
CPU Options: Intel Xeon Scalable or AMD EPYC
Memory: DDR4/DDR5, up to 2TB per server
Storage: High-density NVMe arrays
GPU Support: NVIDIA H100/A100 for AI/ML
Expansion: PCIe Gen4/Gen5 slots

Networking

Architecture: Spine-leaf topology
Switch Type: 2U spine + ToR switches
Port Options: 25/100/400 GbE
InfiniBand: Optional HDR/NDR support
Redundancy: Full mesh redundancy
Management: SDN-ready architecture

Power & Cooling

Power Input: 3-phase 415V AC
Distribution: Dual redundant PDUs
Redundancy: 2N PSU configuration
Cooling Method: Single-phase immersion
Fluid Type: Dielectric mineral oil
Heat Rejection: Dry cooler or liquid loop
PUE: < 1.1 typical

Monitoring & Management

BMC: IPMI 2.0 / Redfish API
Sensors: Temperature, fluid level, power, flow, pressure
Alerts: SNMP, syslog, email, SMS, webhooks
Dashboard: Enterprise management console
Integration: Full DCIM/monitoring API
Analytics: Predictive maintenance & ML insights

Use Cases

AI/ML training clusters, sovereign compute hubs, BFSI core infra, HPC research.

Value Proposition

High-density, compact 1 MW clusters possible with 10 pods.

Configuration

12U Configuration

Compute Pod	Storage Pod	AI Pod
AMD EPYC 9554 20 CPU 1280 CPU cores 20TB DDR5 ECC RAM Fiber 100G 2N Redundant	AMD EPYC 9554 3 Peta-byte storage Fiber 100G Interconnect 2N Redundant	AMD EPYC 9554 1280 CPU cores 20TB DDR5 ECC RAM Fiber 100G Interconnect 15 NVIDIA RTX 5090 1575 FP32 (TFLOPS) 25140 (TOPS, INT8) 12570 (TOPS, FP16/BF16) 480 GB VRAM TDP 15000W

24U Configuration

Compute Pod	Storage Pod	AI Pod
AMD EPYC 9554 40 CPU 2560 CPU cores 40TB DDR5 ECC RAM Fiber 100G 2N Redundant	AMD EPYC 9554 6 Peta-byte storage Fiber 100G Interconnect 2N Redundant	AMD EPYC 9554 2560 CPU cores 40TB DDR5 ECC RAM Fiber 100G Interconnect 30 NVIDIA RTX 5090 3150 FP32 (TFLOPS) 50280 (TOPS, INT8) 25140 (TOPS, FP16/BF16) 960 GB VRAM TDP 30000W

42U Configuration

Compute Pod	Storage Pod	AI Pod
AMD EPYC 9554 80 CPU 5120 CPU cores 80TB DDR5 ECC RAM Fiber 100G 2N Redundant	AMD EPYC 9554 12 Peta-byte storage Fiber 100G Interconnect 2N Redundant	AMD EPYC 9554 5120 CPU cores 80TB DDR5 ECC RAM Fiber 100G Interconnect 60 NVIDIA RTX 5090 6300 FP32 (TFLOPS) 100560 (TOPS, INT8) 50280 (TOPS, FP16/BF16) 1920 GB VRAM TDP 60000W

Common Benefits Across All Pods

Ultra-Low PUE

PUE <1.1 across all configurations, dramatically reducing energy consumption

10× Density

Achieve 10 times higher compute density compared to traditional air-cooled infrastructure

Silent Operation

No fans or pumps required, enabling deployment in noise-sensitive environments

Plug & Play

Pre-configured and tested, ready for immediate deployment upon delivery

100% Made in India

Complete domestic manufacturing supporting local innovation and supply chains

Lower TCO

Reduced operational costs through energy savings, maintenance reduction, and longer hardware life

Complete Range of Immersion-Cooled Pods

Quick Comparison

12U Pod – Compact Edge Pod

Technical Specifications

Detailed Configuration

Server Configuration

Networking

Power & Cooling

Monitoring & Management

Use Cases

Value Proposition

24U Pod – Mid-Range Deployment

Technical Specifications

Detailed Configuration

Server Configuration

Networking

Power & Cooling

Monitoring & Management

Use Cases

Value Proposition

42U Pod – Full Rack Pod

Technical Specifications

Detailed Configuration

Server Configuration

Networking

Power & Cooling

Monitoring & Management

Use Cases

Value Proposition

Configuration

12U Configuration

24U Configuration

42U Configuration

Common Benefits Across All Pods

Ultra-Low PUE

10× Density

Silent Operation

Plug & Play

100% Made in India

Lower TCO

Ready to Deploy Your DCBx Solution?