NVIDIA L40 48GB vs NVIDIA RTX 2000 Ada 16GB

The L40 delivers 7.5x the FP8 throughput of the RTX 2000 Ada (362 vs 48 TFLOPS dense). The L40 also carries 32 GB more memory (48 GB vs 16 GB).

L40: Ada Lovelace, 2023 RTX 2000 Ada: Ada Lovelace, 2024
FP8 dense lead
7.5x
L40: 362 vs 48 TFLOPS
Memory
48 vs 16 GB
32 GB more for the L40
Bandwidth
864 GB/s vs 224 GB/s
L40 moves data faster
TDP
300 W vs 70 W
RTX 2000 Ada draws 230 W less
See rental pricing from $0.24/hr Download comparison sheet

Share this page

NVIDIA L40 48GB vs NVIDIA RTX 2000 Ada 16GB

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

Printable version →

NVIDIA L40 48GB

Leads on FP8 Full specs →

What is the NVIDIA L40?

The L40 is a GPU from NVIDIA on the Ada Lovelace 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. L40 full specifications →

Can I rent the L40 in the cloud?

Yes. Cloud providers tracked by Flopper rent the L40 by the hour, and the pricing page compares their current rates and the price history side by side. L40 rental prices →

How does the L40 compare with other accelerators?

This page sets the L40 against the RTX 2000 Ada. The L40 is the older generation of the two. Both are NVIDIA parts, and both ship as a PCIe card, so they fit the same class of system. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the L40 figures come from?

Every L40 figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

NVIDIA RTX 2000 Ada 16GB

Full specs →

What is the NVIDIA RTX 2000 Ada?

The RTX 2000 Ada is a GPU from NVIDIA on the Ada Lovelace 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 2000 Ada full specifications →

Can I rent the RTX 2000 Ada in the cloud?

Yes. Cloud providers tracked by Flopper rent the RTX 2000 Ada by the hour, and the pricing page compares their current rates and the price history side by side. RTX 2000 Ada rental prices →

How does the RTX 2000 Ada compare with other accelerators?

This page sets the RTX 2000 Ada against the L40. The RTX 2000 Ada is the newer generation of the two. Both are NVIDIA parts, and both ship as a PCIe card, so they fit the same class of system. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the RTX 2000 Ada figures come from?

Every RTX 2000 Ada figure on this page traces to NVIDIA's own datasheet or specification table. How Flopper verifies specifications →

Specifications

L40
RTX 2000 Ada
Architecture
Ada LovelaceAda Lovelace
Launch Year
20232024
Form Factor
PCIePCIe
Memory
48 GB16 GB
Memory Bandwidth
864 GB/s224 GB/s
TDP
300 W70 W
Process Node
4nm4nm

Performance (TFLOPS)

L40
RTX 2000 Ada
No verified data available0.2 TFLOPS
91 TFLOPS 12 TFLOPS
91 TFLOPS
181 TFLOPS sparse
12 TFLOPS
sparse not published
181 TFLOPS
362 TFLOPS sparse
24 TFLOPS
sparse not published
181 TFLOPS
362 TFLOPS sparse
24 TFLOPS
sparse not published
362 TFLOPS
724 TFLOPS sparse
48 TFLOPS
sparse not published
FP6
No verified data available No verified data available
No verified data available No verified data available
362 TOPS
724 TOPS sparse
48 TOPS
sparse not published

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

The RTX 2000 Ada is the newer part: Ada Lovelace (2024) against the L40's Ada Lovelace (2023).

The L40 carries 32 GB more memory (48 GB against 16 GB), so larger models fit on one GPU before you shard. Bandwidth is 864 GB/s against 224 GB/s, which sets the ceiling on decode speed.

Power is 300 W against 70 W, and at FP32 the L40 does more work per watt. Weight the FP8 and FP16 rows for LLM work and the FP64 row for HPC.

New to the terms? The glossary explains FP64, FP32, TF32, BF16, FP16, FP8, INT8, Sparsity and TFLOPS. For how the generations relate, see NVIDIA architectures explained.

Frequently asked questions

Is the L40 faster than the RTX 2000 Ada?

At FP8 precision the L40 reaches 362 TFLOPS dense against 48 TFLOPS for the RTX 2000 Ada. The performance table on this page lists every published precision for both GPUs.

Which should you choose, the L40 or the RTX 2000 Ada?

Choose the L40 for training and high-throughput inference: it has 7.5x the FP8 throughput of the RTX 2000 Ada. It also carries 32 GB more memory, so larger models fit on one GPU before you shard. The L40 draws 230 W more power, so budget for cooling and power delivery before you commit.

Which has more memory, the L40 or the RTX 2000 Ada?

The L40 carries 48 GB of memory versus 16 GB for the RTX 2000 Ada. Memory bandwidth is 864 GB/s for the L40 and 224 GB/s for the RTX 2000 Ada.

How much power do the L40 and the RTX 2000 Ada draw?

The L40 is rated at 300 W TDP and the RTX 2000 Ada at 70 W. On FP32 throughput per watt, the L40 is the more efficient part.

How much do the L40 and the RTX 2000 Ada cost to rent?

Live cloud listings tracked by Flopper price the L40 from $0.33 per GPU-hour across 9 providers, and the RTX 2000 Ada from $0.24 per GPU-hour across 2 providers. The rental section on this page links to every provider, rate and the price history for each GPU.

Where to Rent

NVIDIA L40 48GB

Rent from

$0.33 / GPU-hr

Lowest of 18 live listings across 9 providers · median $0.88

See all L40 rental prices →

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

NVIDIA RTX 2000 Ada 16GB

Rent from

$0.24 / GPU-hr

Lowest of 3 live listings across 2 providers

See all RTX 2000 Ada 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