NVIDIA A100 PCIe 40GB vs NVIDIA RTX A6000 48GB

The A100 delivers 2.0x the FP16 throughput of the RTX A6000 (312 vs 155 TFLOPS dense). The RTX A6000 counters with 8 GB more memory (48 GB vs 40 GB).

A100: Ampere, 2020 RTX A6000: Ampere, 2020
FP16 dense lead
2.0x
A100: 312 vs 155 TFLOPS
Memory
40 vs 48 GB
8 GB more for the RTX A6000
Bandwidth
1.6 TB/s vs 768 GB/s
A100 moves data faster
TDP
250 W vs 300 W
A100 draws 50 W less
See rental pricing from $0.33/hr Download comparison sheet

Share this page

NVIDIA A100 PCIe 40GB vs NVIDIA RTX A6000 48GB

A post with this link shows the card on the Image tab as its preview.

Printable version →

NVIDIA A100 PCIe 40GB

Leads on FP16 Full specs →

What is the NVIDIA A100?

The A100 is a GPU from NVIDIA on the Ampere architecture, an accelerator for AI training, inference and high-performance computing. It is a PCIe add-in card, so it fits standard servers and workstations. A100 full specifications →

Can I rent the A100 in the cloud?

Yes. Cloud providers tracked by Flopper rent the A100 by the hour, and the pricing page compares their current rates and the price history side by side. A100 rental prices →

How does the A100 compare with other accelerators?

This page sets the A100 against the RTX A6000. Both parts launched in the same year. Both are NVIDIA parts, and both ship as a PCIe card, so they fit the same class of system. Flopper also compares it head to head with the A100 SXM4 80GB, across every published precision, memory and power. A100 vs A100 SXM4 80GB →

Where do the A100 figures come from?

Every A100 figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

NVIDIA RTX A6000 48GB

Full specs →

What is the NVIDIA RTX A6000?

The RTX A6000 is a GPU from NVIDIA on the Ampere architecture, an accelerator for AI training, inference and high-performance computing. It is a PCIe add-in card, so it fits standard servers and workstations. RTX A6000 full specifications →

Can I rent the RTX A6000 in the cloud?

Yes. Cloud providers tracked by Flopper rent the RTX A6000 by the hour, and the pricing page compares their current rates and the price history side by side. RTX A6000 rental prices →

How does the RTX A6000 compare with other accelerators?

This page sets the RTX A6000 against the A100. Both parts launched in the same year. Both are NVIDIA parts, and both ship as a PCIe card, so they fit the same class of system. Flopper also compares it head to head with the SC11 FP300, across every published precision, memory and power. RTX A6000 vs SC11 FP300 →

Where do the RTX A6000 figures come from?

Every RTX A6000 figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

Specifications

A100
RTX A6000
Architecture
AmpereAmpere
Launch Year
20202020
Form Factor
PCIePCIe
Memory
40 GB48 GB
Memory Bandwidth
1.6 TB/s768 GB/s
TDP
250 W300 W
Process Node
7nm8nm

Performance (TFLOPS)

A100
RTX A6000
9.7 TFLOPS 0.6 TFLOPS
20 TFLOPS 39 TFLOPS
156 TFLOPS
312 TFLOPS sparse
77 TFLOPS
155 TFLOPS sparse
312 TFLOPS
624 TFLOPS sparse
155 TFLOPS
310 TFLOPS sparse
312 TFLOPS
624 TFLOPS sparse
155 TFLOPS
310 TFLOPS sparse
No verified data available No verified data available
FP6
No verified data available No verified data available
No verified data available No verified data available
624 TOPS
1,248 TOPS sparse
310 TOPS
619 TOPS sparse

Figures are dense. Where a vendor publishes a figure it sits under the dense one and never decides a winner. Hover any precision for its definition, or read the GPU glossary.

FLOPS by Precision

What actually differs

Choose by the limit your job hits first, because here the workstation card holds more. The RTX A6000’s extra 8 GB decides whether a model, its KV cache and its optimizer state fit on one card. The A100 PCIe 40GB reads its smaller memory at 1,555 GB/s against 768 GB/s, so anything that fits trains and decodes roughly twice as fast.

The rest of the spec sheet splits rather than tilts. Simulation codes belong on the A100, which runs 9.7 TFLOPS FP64 where the A6000 barely runs it, while plain FP32 rendering favours the A6000. MIG partitions the A100 for a shared server; the A6000 trades that for a blower and display outputs, so it runs in a tower where the passive A100 overheats.

The trap is supply. The A6000 rents from $0.30 per GPU-hour on its pricing page, while Flopper tracks a single A100 40GB listing.

New to the terms? The glossary explains FP64, FP32, TF32, BF16, FP16, INT8, Sparsity and TFLOPS. For how the generations relate, see NVIDIA architectures explained.

Frequently asked questions

Is the A100 40GB faster than the RTX A6000?

On AI work that fits in 40 GB, yes. The A100 PCIe 40GB moves 1,555 GB/s against 768 GB/s and runs 312 TFLOPS BF16 dense against 154.8 TFLOPS, so training steps and token generation run roughly twice as fast. The RTX A6000 leads only on plain FP32, at 38.71 TFLOPS against 19.5 TFLOPS.

Should I choose the A100 40GB or the RTX A6000?

Choose the A100 for bandwidth-bound inference, FP64 simulation and shared servers that need MIG partitioning. Choose the RTX A6000 when the extra 8 GB decides whether a model fits on one card, or when the GPU lives in a desk-side tower that needs a blower fan and display outputs.

Does the RTX A6000 have more memory than the A100?

Yes. The RTX A6000 carries 48 GB of GDDR6 against 40 GB of HBM2 on the A100 PCIe 40GB. The A100's memory is smaller but twice as fast, 1,555 GB/s against 768 GB/s, so the A6000 holds more and the A100 reads it faster.

Can you rent an A100 40GB or an RTX A6000?

Flopper's listings show 23 RTX A6000 listings across 14 providers, from $0.30 to $1.93 per GPU-hour, with the lowest rate on spot capacity. The A100 PCIe 40GB has one tracked listing, on Vast.ai at $0.60 per GPU-hour on community capacity.

Where to Rent

NVIDIA A100 PCIe 40GB

Rent from

$0.53 / GPU-hr

Lowest of 1 live listing across 1 provider

See all A100 rental prices →

Price history and every provider on the pricing page. Some links are affiliate links.

NVIDIA RTX A6000 48GB

Rent from

$0.33 / GPU-hr

Lowest of 23 live listings across 14 providers · median $0.55

See all RTX A6000 rental prices →

Price history and every provider on the pricing page. Some links are affiliate links.

Get Comparison Updates

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

© 2026 Flopper.io - Compare the GPUs Powering AI