NVIDIA RTX A4000 16GB vs NVIDIA A100 SXM4 40GB

The A100 delivers 16.3x the FP16 throughput of the RTX A4000 (312 vs 19 TFLOPS dense). The A100 also carries 24 GB more memory (40 GB vs 16 GB).

RTX A4000: Ampere, 2021 A100: Ampere, 2020
FP16 dense lead
16.3x
A100: 312 vs 19 TFLOPS
Memory
16 vs 40 GB
24 GB more for the A100
Bandwidth
448 GB/s vs 1.6 TB/s
A100 moves data faster
TDP
140 W vs 400 W
RTX A4000 draws 260 W less

Specifications

RTX A4000
A100
Architecture
AmpereAmpere
Launch Year
20212020
Form Factor
PCIeSXM
Memory
16 GB40 GB
Memory Bandwidth
448 GB/s1.6 TB/s
TDP
140 W400 W
Process Node
8nm7nm

Performance (TFLOPS)

RTX A4000
A100
FP64
0.3 TFLOPS 9.7 TFLOPS
FP32
19 TFLOPS 20 TFLOPS
TF32
19 TFLOPS
sparse not published
156 TFLOPS
312 TFLOPS sparse
BF16
19 TFLOPS
sparse not published
312 TFLOPS
624 TFLOPS sparse
FP16
19 TFLOPS
sparse not published
312 TFLOPS
624 TFLOPS sparse
FP8
No verified data available No verified data available
FP6
No verified data available No verified data available
FP4
No verified data available No verified data available
INT8
38 TOPS
sparse not published
624 TOPS
1,248 TOPS sparse

FLOPS by Precision

What actually differs

The RTX A4000 is the newer part: Ampere, launched in 2021, against the A100's Ampere from 2020. 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 RTX A4000 against the A100: FP64 0.3 vs 9.7 TFLOPS, FP32 19 vs 20 TFLOPS, FP16 19 vs 312 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 16 GB against 40 GB, fed at 448 GB/s versus 1.6 TB/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 140 W for the RTX A4000 and 400 W for the A100. At FP32 that works out to 0.14 against 0.05 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 RTX A4000 faster than the A100?

At FP16 precision the A100 reaches 312 TFLOPS dense against 19 TFLOPS for the RTX A4000. The performance table on this page lists every published precision for both GPUs.

Which has more memory, the RTX A4000 or the A100?

The A100 carries 40 GB of memory versus 16 GB for the RTX A4000. Memory bandwidth is 448 GB/s for the RTX A4000 and 1.6 TB/s for the A100.

How much power do the RTX A4000 and the A100 draw?

The RTX A4000 is rated at 140 W TDP and the A100 at 400 W. On FP32 throughput per watt, the RTX A4000 is the more efficient part.

Can I rent the RTX A4000 or the A100 in the cloud?

Yes. Live cloud listings tracked by Flopper start at $0.15 per GPU hour. The rental pricing section on this page lists current providers and rates for both GPUs.

Where to Rent

NVIDIA RTX A4000 16GB

ProviderConfigurationPrice/GPU-hrChecked
Hyperstack
1× RTX A4000 16GB $0.15 2h ago View →
RunPod
1× RTX A4000 16GB Community $0.17 2h ago View →
RunPod
1× RTX A4000 16GB $0.25 2h ago View →
Paperspace
2× RTX A4000 16GB $0.80 2h ago View →
All RTX A4000 listings and price history →

NVIDIA A100 SXM4 40GB

ProviderConfigurationPrice/GPU-hrChecked
DataCrunch
1× A100 SXM4 Spot $0.65 2h ago View →
RunPod
1× A100 SXM4 Community $1.00 2h ago View →
DataCrunch
1× A100 SXM4 $1.29 2h ago View →
Microsoft Azure
8× A100 SXM4 $3.40 8h ago View →
All A100 listings and price history →

Get Comparison Updates

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