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

Ampere

NVIDIA RTX A6000

wins

1

The RTX A6000 is NVIDIA's Ampere professional workstation GPU with 48GB GDDR6 and NVLink support.

Professional workstation AILarge model inferenceRendering

Performance comparison

Visual bars — winner highlighted

NVIDIA GB300 NVL72

Metric

NVIDIA RTX A6000

FP16 Tensor TFLOPS (dense)

7.5k T
155 T

Bandwidth

16.0k GB/s
768 GB/s

VRAM

288 GB
48 GB

TDP (lower=better)

1.2k W
300 W

TFLOPS/watt

6.25
0.52

GB/s per watt

13.33
2.56

Full specification table

SpecNVIDIA GB300 NVL72NVIDIA RTX A6000
ArchitectureBlackwell UltraAmpere
VRAM288 GB HBM3e (per GPU)48 GB GDDR6
VRAM typeHBM3e (per GPU)GDDR6
Memory bandwidth16,000 GB/s768 GB/s
FP16 Tensor TFLOPS (dense)7,500 TFLOPS154.8 TFLOPS
TDP1,200 W300 W
TFLOPS/watt6.25 T/W0.52 T/W
GB/s per watt13.33 GB/s/W2.56 GB/s/W
NVLinkYesYes
Release year20262020

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 RTX A6000

FP16 TFLOPS154.8 TFLOPS
TDP300 W
TFLOPS/watt0.516 T/W
GB/s per watt2.56 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 RTX A6000 for:

  • Professional workstation AI
  • Large model inference
  • Rendering

Weaknesses / Limitations

  • Older Ampere architecture — lower TFLOPS than Ada equivalents at same VRAM
  • GDDR6 bandwidth (768 GB/s) lower than HBM-based GPUs
  • Superseded by RTX 6000 Ada and RTX PRO 6000 on new deployments