NVIDIA H100 PCIe 80GB Tensor Core GPU; seller PCIe variant
80 GB HBM2e with ECC · PCIeGen5 x16
New
Ready stock
₹38,56,041.24
Per unit, incl. GST
NVIDIA H100 GPU servers
Compare exact GPU and platform offers. Complete systems are identified separately from cards and modules.
FROM THE HARDWARE CATALOGUE
Purchase options in this GPU family.
Compare the exact listed card, module, baseboard or complete system before ordering.
Dell H100 NVL Tensor Core GPU 94GB Memory Interface 6016 Bit HBM3 Memory Bandwidth 3938GB/s PCI-E 5.0 x16
Manufacturer part number 330-BCFX
94 GB HBM3 · PCIe Gen5 x16 / NVLink
New
Ready stock
₹44,51,550
Per unit, incl. GST
NVIDIA H100 SXM5 80GB HBM3 Tensor Core GPU module; one module for HGX H100 baseboard
80 GB HBM3 · SXM / NVLink; host PCIe Gen5 interconnect
New
Ready stock
₹28,08,400
Per unit, incl. GST
Compare the exact offered unit: a GPU card, module, baseboard and complete system have different supply scopes. Listed purchase prices include 18% GST. International delivery and commercial terms are confirmed by quotation.
Explore all models and part numbers ↗CONFIGURATION & QUOTATIONSRequest a hardware quoteGet pricing and availability for your model, part number or configuration.Quote form
Tell us what you need
Our team will confirm the exact configuration, condition, delivery and warranty terms in your quotation.
View NVIDIA model specifications and reference images ↗
CONFIGURATION FIRST
Build around the application.
Configure an H100 system for AI training, LLM inference, fine-tuning and high-performance computing.
Discuss H100 PCIe, H100 NVL or an integrated H100 SXM/HGX system. These are distinct configurations with different platform, interconnect, power and cooling requirements. We define the complete server around your application and scale.
NVIDIA Hopper
NVIDIA H100 GPU servers
NVIDIA H100 PCIe
PCIe server configurations, selected around your workload and platform requirements.
NVIDIA H100 NVL
Enterprise PCIe configurations with supported NVLink connectivity.
NVIDIA H100 SXM
Integrated HGX platforms for demanding multi-GPU workloads.
H100, H200 and H20 are distinct products. PCIe/NVL and SXM/HGX configurations need different server designs; the quotation identifies the actual GPU, interconnect and host.
THE COMPLETE BUILD
Every component has a role.
A GPU choice is the starting point. Configure a balanced system around the application.
GPU & platform
Select the model, memory requirements, GPU count and supported PCIe or integrated platform.
CPU & system memory
Match processor capacity, RAM and PCIe resources to your compute and data pipeline.
Storage & networking
Plan NVMe capacity, data throughput, network connectivity and multi-node communication.
Chassis, power & cooling
Review dimensions, rack space, power supplies and thermal requirements for the exact build.
ENTERPRISE PLATFORMS.
YOUR CHOICE OF BUILD.
SupermicroDell PowerEdgeHPECustom multi-GPU chassis
LET'S CONFIGURE YOUR NEXT SYSTEM
Start with your workload.
We'll talk hardware.
Tell us what you need to run, where you need it and your budget.