Compute Comparison
vs
All comparisons →

Hopper

NVIDIA H200 SXM 141GB

wins

4

The H200 SXM is the SXM5 form factor of the H200, with 141 GB HBM3e and 4.8 TB/s bandwidth.

Very large LLMsMemory-bound inferenceMulti-modal models

Ada Lovelace

NVIDIA RTX 4060 8GB

wins

1

The RTX 4060 is the most power-efficient Ada Lovelace GPU at 115W TDP. 8GB GDDR6 limits it to small quantized models (up to ~7B at INT4).

Ultra-budget inferenceSmall model experimentationEntry-level Ada AI

Performance comparison

Visual bars — winner highlighted

NVIDIA H200 SXM 141GB

Metric

NVIDIA RTX 4060 8GB

FP16 Tensor TFLOPS (dense)

989 T
30 T

Bandwidth

4.8k GB/s
272 GB/s

VRAM

141 GB
8 GB

TDP (lower=better)

700 W
115 W

TFLOPS/watt

1.41
0.26

GB/s per watt

6.86
2.37

Full specification table

SpecNVIDIA H200 SXM 141GBNVIDIA RTX 4060 8GB
ArchitectureHopperAda Lovelace
VRAM141 GB HBM3e8 GB GDDR6
VRAM typeHBM3eGDDR6
Memory bandwidth4,800 GB/s272 GB/s
FP16 Tensor TFLOPS (dense)989 TFLOPS30.2 TFLOPS
TDP700 W115 W
TFLOPS/watt1.41 T/W0.26 T/W
GB/s per watt6.86 GB/s/W2.37 GB/s/W
NVLinkYesNo
Release year20242023

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 H200 SXM 141GB

FP16 TFLOPS989 TFLOPS
TDP700 W
TFLOPS/watt1.413 T/W
GB/s per watt6.86 GB/s/W

NVIDIA RTX 4060 8GB

FP16 TFLOPS30.2 TFLOPS
TDP115 W
TFLOPS/watt0.263 T/W
GB/s per watt2.37 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 H200 SXM 141GB for:

  • Very large LLMs
  • Memory-bound inference
  • Multi-modal models

Weaknesses / Limitations

  • Very limited cloud availability — fewer providers than H100
  • Premium pricing over H100 with same FP16 TFLOPS
  • 700 W TDP — same power envelope as H100 with no efficiency gain

Choose NVIDIA RTX 4060 8GB for:

  • Ultra-budget inference
  • Small model experimentation
  • Entry-level Ada AI

Weaknesses / Limitations

  • 8 GB VRAM — only fits quantized models up to ~7B parameters
  • Very low bandwidth (272 GB/s) — severe bottleneck for inference throughput
  • Not viable for fine-tuning beyond tiny models