Compute Comparison
vs
All comparisons →

Blackwell Ultra

NVIDIA GB300 NVL72

wins

4

The GB300 NVL72 is NVIDIA's Blackwell Ultra rack-scale system — 72 GPUs in a single rack delivering 7,500 TFLOPS FP16 per GPU.

Next-gen frontier trainingTrillion-param inferenceRack-scale AI

Hopper

NVIDIA H100 PCIe 80GB

wins

1

The H100 PCIe is the more affordable form factor of the H100.

Budget Hopper inferencePCIe server deploymentsCost-sensitive LLM serving

Performance comparison

Visual bars — winner highlighted

NVIDIA GB300 NVL72

Metric

NVIDIA H100 PCIe 80GB

FP16 Tensor TFLOPS (dense)

7.5k T
756 T

Bandwidth

16.0k GB/s
2.0k GB/s

VRAM

288 GB
80 GB

TDP (lower=better)

1.2k W
350 W

TFLOPS/watt

6.25
2.16

GB/s per watt

13.33
5.71

Full specification table

SpecNVIDIA GB300 NVL72NVIDIA H100 PCIe 80GB
ArchitectureBlackwell UltraHopper
VRAM288 GB HBM3e (per GPU)80 GB HBM2e
VRAM typeHBM3e (per GPU)HBM2e
Memory bandwidth16,000 GB/s2,000 GB/s
FP16 Tensor TFLOPS (dense)7,500 TFLOPS756 TFLOPS
TDP1,200 W350 W
TFLOPS/watt6.25 T/W2.16 T/W
GB/s per watt13.33 GB/s/W5.71 GB/s/W
NVLinkYesNo
Release year20262022

Live cloud pricing

On-demand pricing across providers — updated in real time

Per hourPer GB VRAM/hr
Loading prices…

Power efficiency analysis

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

NVIDIA GB300 NVL72

FP16 TFLOPS7,500 TFLOPS
TDP1,200 W
TFLOPS/watt6.250 T/W
GB/s per watt13.33 GB/s/W

NVIDIA H100 PCIe 80GB

FP16 TFLOPS756 TFLOPS
TDP350 W
TFLOPS/watt2.160 T/W
GB/s per watt5.71 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 GB300 NVL72 for:

  • Next-gen frontier training
  • Trillion-param inference
  • Rack-scale AI

Weaknesses / Limitations

  • Rack-scale system — not a single GPU, requires full NVL72 rack deployment
  • Extremely high cost — rack-scale pricing only
  • Very limited availability, primarily hyperscaler deployments

Choose NVIDIA H100 PCIe 80GB for:

  • Budget Hopper inference
  • PCIe server deployments
  • Cost-sensitive LLM serving

Weaknesses / Limitations

  • Lower FP16 TFLOPS (756) and bandwidth (2,000 GB/s) than H100 SXM5
  • No NVLink — cannot scale across cards for very large models
  • Higher cost than A100 with only moderate performance advantage