
NVIDIA
NVIDIA Jetson AGX Thor Developer Kit
- GPU model
- 2,560-core NVIDIA Blackwell GPU, fifth-gen Tensor Cores, MIG with 10 TPCs
₹8,78,524
Backorder

Level 2
AI, GPU, HPC and enterprise servers, multi-GPU and scientific workstations, storage and networking systems, edge AI and industrial computing.
Servers built around AI accelerators for training and inference, typically 4 or 8 GPUs with high-bandwidth GPU interconnects.
4 products
Rack servers with multiple PCIe or SXM/OAM GPUs for AI, rendering, simulation and virtual desktop workloads.
5 products
Dense compute nodes optimised for memory bandwidth, core count and low-latency interconnects.
6 products

General-purpose rack and tower servers for virtualisation, databases and business applications.
3 products
Deskside workstations with two to four professional GPUs for model development, simulation and rendering.
3 products
Workstations configured for local AI development, fine-tuning and inference.
4 products
Workstations with high core counts, large ECC memory and double-precision-capable compute for research codes.
4 products
Certified workstations for CAD, CAE, CFD and FEA applications.
4 products
Workstations for content creation, data analysis and software development.
3 products
High-capacity servers and JBOD/JBOF enclosures for file, block and object storage.
2 products
Data-centre switches, routers and fabric systems for Ethernet and InfiniBand networks.
2 products
Compact systems with GPUs or NPUs for on-site inference — vision, speech and sensor analytics.
5 products
Embedded and on-robot compute platforms for perception, planning and control.
3 products
Fanless and ruggedised PCs and servers rated for wide temperature ranges, vibration and continuous operation.
3 products
Systems configured for physics, engineering and Monte-Carlo simulation workloads.
5 products
Compute, visualisation and data infrastructure for real-time digital twins of plants, products and facilities.
3 products
Hypervisor hosts, shared storage and networking for consolidated virtual machine estates.
4 products
Integrated compute, storage and network systems for running cloud services on premises.
4 products
In the catalogue

NVIDIA
₹8,78,524
Backorder

Apple
₹6,29,900
Backorder

Supermicro
₹1,61,15,171
Built to order

Supermicro
₹5,12,07,811
Built to order

Supermicro
₹9,03,53,780
Built to order

NVIDIA
₹1,67,47,713
Built to order

NVIDIA
₹11,10,337
Backorder

Size it first
LLM Inference Calculator
GPU memory, GPU count, KV cache and theoretical token throughput for serving a large language model.
AI Training Compute Calculator
Training FLOPs, time to train, GPU-hours, energy and cost for a transformer model.
GPU Memory (VRAM) Calculator
Per-GPU memory for inference, full training, LoRA and QLoRA — with tensor, pipeline and ZeRO parallelism.
AI Infrastructure & Cluster Planner
From GPU count to servers, racks, power, cooling, network ports, storage and energy cost.
System Power & PSU Calculator
Sustained and peak power, PSU size, wall power and running cost for a workstation or server.
Rack Power & Density Calculator
Rack kW, current per feed, PDU sizing and heat for a rack of servers.
HPC Cluster Planner
Peak and estimated sustained FLOPS, power, heat and racks for a CPU/GPU cluster.
Memory Bandwidth & Channel Calculator
Theoretical memory bandwidth, capacity and bandwidth per core from channels, speed and population.