Compute Comparison
vs
All comparisons →

Blackwell

NVIDIA B200 180GB

wins

2

The B200 is NVIDIA's Blackwell flagship — 4.5 PFLOPS FP16 and 8 TB/s memory bandwidth make it the most powerful GPU available for AI training at scale.

Next-gen LLM trainingTrillion-parameter modelsAI factories

Blackwell

NVIDIA GB200 NVL72

wins

3

The GB200 NVL72 is NVIDIA's most powerful AI system.

Trillion-parameter trainingAI factory scaleNext-gen LLM inference

Performance comparison

Visual bars — winner highlighted

NVIDIA B200 180GB

Metric

NVIDIA GB200 NVL72

FP16 TFLOPS

4.5k T
9.0k T

Bandwidth

8.0k GB/s
16.0k GB/s

VRAM

1.8k GB
3.8k GB

TDP (lower=better)

1.0k W
2.7k W

TFLOPS/watt

4.50
3.33

GB/s per watt

8.00
5.93

Full specification table

SpecNVIDIA B200 180GBNVIDIA GB200 NVL72
ArchitectureBlackwellBlackwell
VRAM180 GB HBM3e384 GB HBM3e (per GPU)
VRAM typeHBM3eHBM3e (per GPU)
Memory bandwidth8,000 GB/s16,000 GB/s
FP16 TFLOPS4,500 TFLOPS9,000 TFLOPS
TDP1,000 W2,700 W
TFLOPS/watt4.50 T/W3.33 T/W
GB/s per watt8.00 GB/s/W5.93 GB/s/W
NVLinkYesYes
Release year20242025

Live cloud pricing

On-demand $/hr across providers — updated in real time

Loading prices…

Power efficiency analysis

TFLOPS/watt and GB/s/watt — critical for data center TCO

NVIDIA B200 180GB

FP16 TFLOPS4,500 TFLOPS
TDP1,000 W
TFLOPS/watt4.500 T/W
GB/s per watt8.00 GB/s/W

NVIDIA GB200 NVL72

FP16 TFLOPS9,000 TFLOPS
TDP2,700 W
TFLOPS/watt3.333 T/W
GB/s per watt5.93 GB/s/W

TFLOPS/watt measures compute efficiency — how much AI throughput you get per watt of power consumed. For data centers with PUE of 1.2–1.5, a 10% improvement in TFLOPS/watt translates directly to lower electricity costs and cooling requirements. GB/s/watt measures memory bandwidth efficiency, which is the binding constraint for memory-bound LLM inference workloads.

When to choose each GPU

Choose NVIDIA B200 180GB for:

  • Next-gen LLM training
  • Trillion-parameter models
  • AI factories

Choose NVIDIA GB200 NVL72 for:

  • Trillion-parameter training
  • AI factory scale
  • Next-gen LLM inference