NVIDIA A10G 24GB vs NVIDIA A40 48GB

The A40 delivers 2.1x the FP16 throughput of the A10G (150 vs 70 TFLOPS dense). The A40 also carries 24 GB more memory (48 GB vs 24 GB).

A10G: Ampere, 2022 A40: Ampere, 2021
FP16 dense lead
2.1x
A40: 150 vs 70 TFLOPS
VRAM
24 vs 48 GB
24 GB more for the A40
Bandwidth
600 GB/s vs 696 GB/s
A40 moves data faster
TDP
300 W vs 300 W
Same power budget

Specifications

A10G
A40
Architecture
AmpereAmpere
Launch Year
20222021
Form Factor
PCIePCIe
VRAM
24 GB48 GB
Memory Bandwidth
600 GB/s696 GB/s
TDP
300 W300 W
Process Node
8nm

Performance (TFLOPS)

A10G
A40
FP32
35.0 TFlops 37.4 TFlops
FP16
70.0 TFlops
sparse not published
149.7 TFlops
299.4 TFLOPS sparse
BF16
70.0 TFlops
sparse not published
149.7 TFlops
299.4 TFLOPS sparse
INT8
140.0 TFlops
sparse not published
299.3 TFlops
598.6 TOPS sparse

FLOPS by Precision

What actually differs

The A10G is the newer part: Ampere, launched in 2022, against the A40's Ampere from 2021. Newer architectures typically add lower-precision formats and better throughput per watt, so check the precision rows your workload actually uses.

Dense throughput for the A10G against the A40: FP32 35 vs 37 TFLOPS, FP16 70 vs 150 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 24 GB against 48 GB, fed at 600 GB/s versus 696 GB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.

Power budgets are 300 W for the A10G and 300 W for the A40. At FP32 that works out to 0.12 against 0.12 TFLOPS per watt.

For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.

Frequently asked questions

Is the A10G faster than the A40?

At FP16 precision the A40 reaches 150 TFLOPS dense against 70 TFLOPS for the A10G. The performance table on this page lists every published precision for both GPUs.

Which has more memory, the A10G or the A40?

The A40 carries 48 GB of VRAM versus 24 GB for the A10G. Memory bandwidth is 600 GB/s for the A10G and 696 GB/s for the A40.

How much power do the A10G and the A40 draw?

The A10G is rated at 300 W TDP and the A40 at 300 W. On FP32 throughput per watt, the A40 is the more efficient part.

Can I rent the A10G or the A40 in the cloud?

Yes. Live cloud listings tracked by Flopper start at $0.35 per GPU hour. The rental pricing section on this page lists current providers and rates for both GPUs.

Where to Rent

NVIDIA A40 48GB

ProviderConfigurationPrice/GPU-hrChecked
RunPod
1× A40 48GB Community $0.35 just now View →
RunPod
1× A40 48GB $0.44 just now View →
All A40 listings and price history →

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.