Compute Comparison
vs
All comparisons →

Ampere

NVIDIA A10G

wins

1

The NVIDIA A10G is the AWS-specific variant of the A10, available on g5 instances.

AWS inferenceGraphics workloadsMulti-tenant AI

Ada Lovelace

NVIDIA L4

wins

3

The NVIDIA L4 is the most power-efficient Ada Lovelace GPU at just 72W TDP.

Efficient inferenceVideo transcodingEdge AI

Performance comparison

Visual bars — winner highlighted

NVIDIA A10G

Metric

NVIDIA L4

FP16 TFLOPS

125 T
242 T

Bandwidth

600 GB/s
300 GB/s

VRAM

246 GB
246 GB

TDP (lower=better)

150 W
72 W

TFLOPS/watt

0.83
3.36

GB/s per watt

4.00
4.17

Full specification table

SpecNVIDIA A10GNVIDIA L4
ArchitectureAmpereAda Lovelace
VRAM24 GB GDDR624 GB GDDR6
VRAM typeGDDR6GDDR6
Memory bandwidth600 GB/s300 GB/s
FP16 TFLOPS125 TFLOPS242 TFLOPS
TDP150 W72 W
TFLOPS/watt0.83 T/W3.36 T/W
GB/s per watt4.00 GB/s/W4.17 GB/s/W
NVLinkNoNo
Release year20212023

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 A10G

FP16 TFLOPS125 TFLOPS
TDP150 W
TFLOPS/watt0.833 T/W
GB/s per watt4.00 GB/s/W

NVIDIA L4

FP16 TFLOPS242 TFLOPS
TDP72 W
TFLOPS/watt3.361 T/W
GB/s per watt4.17 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 A10G for:

  • AWS inference
  • Graphics workloads
  • Multi-tenant AI

Choose NVIDIA L4 for:

  • Efficient inference
  • Video transcoding
  • Edge AI