NVIDIA DGX SuperPOD GB200

pod
2304× B200
NVLink, NVSwitch, InfiniBand
2025
FP64
92.2
PFLOPS
FP32
184.3
PFLOPS
Power
3840.0
kW Total
Memory
GB Total

32-rack SuperPOD configuration with GB200 NVL72 systems

We will point you at suppliers who have it. Free, and no signup.

FP64
92.16
PFLOPS
27.0 TFLOPS/kW
FP32
184.32
PFLOPS
54.0 TFLOPS/kW
TF32
2880.00
PFLOPS
1500.0 TFLOPS/kW
FP16
5760.00
PFLOPS
2700.0 TFLOPS/kW

System Details

GPU Configuration

GPU Count: 2304 GPUs
Architecture: Blackwell
Interconnect: NVLink, NVSwitch, InfiniBand

System Specifications

Form Factor: pod
Total Power: 3840.0 kW
Total Memory:
Memory Bandwidth: 17.7 PB/s

Precision Performance Breakdown

PrecisionSystem PerformancePer GPUEfficiency
92.160 PFLOPS 40.0 TFLOPS 27.0 TFLOPS/kW
184.320 PFLOPS 80.0 TFLOPS 54.0 TFLOPS/kW
2880.000 PFLOPS 1250.0 TFLOPS 1500.0 TFLOPS/kW
5760.000 PFLOPS 2500.0 TFLOPS 2700.0 TFLOPS/kW
5760.000 PFLOPS 2500.0 TFLOPS 2700.0 TFLOPS/kW
11520.000 PFLOPS 5000.0 TFLOPS 5400.0 TFLOPS/kW
23040.000 PFLOPS 10000.0 TFLOPS 10800.0 TFLOPS/kW
11520.000 PFLOPS 5000.0 TFLOPS 5400.0 TFLOPS/kW

Where a vendor states a basis, the figure shown is the dense one, and any figure it publishes is listed separately. Vendors commonly headline the sparse number instead: for NVIDIA's 2:4 structured sparsity that is exactly twice the dense figure, though other vendors' sparse modes do not all follow that ratio.

Powered by NVIDIA B200

This system utilizes 2304 × NVIDIA B200 SXM 180GB GPUs, each delivering exceptional performance for AI and HPC workloads.

Per GPU TDP

1000W

Per GPU Memory

180 GB

Process Node

4nm

Architecture

Blackwell

Documentation & Resources

Flopper Spec Sheet

NVIDIA DGX SuperPOD GB200 specifications, generated from our database. Printable.

Vendor product page

NVIDIA DGX SuperPOD GB200 documentation

View ↗

workstation-datasheet-dgx-spark-gtc25-spring-nvidia-us-3716899-web.pdf

NVIDIA • 2025-10-16

View Document ↗

NVIDIA DGX Documentation

System guides, deployment resources

Browse ↗

Typical Use Cases

AI/ML Training
High-Performance Computing
Data Analytics

The NVIDIA DGX SuperPOD GB200 runs 2304× B200 GPUs over NVLink, NVSwitch, InfiniBand, delivering 11.52 EFLOPS FP8 dense.