NVIDIA L40
NVIDIA: specifications, comparisons and alternatives.
The NVIDIA L40 is a graphics card from NVIDIA. According to the manufacturer, the specifications include fp32 compute 90.5 TFLOPS, fp16/bf16 181.05 TFLOPS, fp8 362 TFLOPS. On this page you see the specifications with sources, the Clouwino points and a comparison with similar products, checked on 19 October 2026.
NVIDIA
- 48 GB
- Memory
- GDDR6 (ECC)
- Memory type
- 0,86 TB/s
- Memory bandwidth
- 300 W
- Power (board power or TDP)
You can see the current price and availability with the partner as soon as a link has been set up.
How are the points calculated?
Clouwino points are an arithmetic summary of the specifications the manufacturer states, on a scale of 0 to 100. It is not a benchmark and not a judgement on quality or suitability: a product with fewer points may suit your situation better. Products are compared with other products of the same type. For each property we count how far a product sits between the lowest and highest value in the group:
- FP32 compute (for: consumer), carries heavy weight
- Memory (for: consumer), carries heavy weight
- Memory bandwidth (for: consumer), carries heavy weight
- Compute units (for: consumer), carries light weight
- Boost clock (for: consumer), carries light weight
- Memory (for: professional and data centre), carries medium weight
- Memory bandwidth (for: professional and data centre), carries medium weight
- FP16/BF16 (for: professional and data centre), carries heavy weight
- FP8 (for: professional and data centre), carries medium weight
- FP64 (for: professional and data centre), carries medium weight
If too much data is missing, we do not show points. Points are relative: when a new product is added, they can shift. Read more at how we compare.
Specifications of NVIDIA L40
All specifications the manufacturer lists, per item. Want to see another product next to it? Pick one below.
| Specificatie | NVIDIA L40 |
|---|---|
| Overview | |
| Clouwino points | 18 punten |
| General | |
| Architecture | Ada Lovelace |
| Manufacturing process | Not stated by the manufacturer |
| Form | PCIe, dual-slot (passief) |
| Launch | Not stated by the manufacturer |
| Compute | |
| FP32 compute | 90,5 TFLOPS |
| FP64 | Not stated by the manufacturer |
| FP16/BF16 | 181,05 TFLOPS |
| FP8 | 362 TFLOPS |
| FP4 | Not stated by the manufacturer |
| Performance notes | From the NVIDIA L40 datasheet (January 2023, 'preliminary, subject to change'), dense labelled: FP16/BF16 Tensor Core 181.05 dense (362.1 with sparsity); FP8 362 dense (724 with sparsity); TF32 90.5 dense (181 sparse); INT8 362 TOPS dense (724 sparse). FP32 90.5 TFLOPS without a label. FP64 and FP4 not stated. |
| Memory | |
| Memory | 48 GB |
| Memory type | GDDR6 (ECC) |
| Memory bandwidth | 0,86 TB/s |
| Power and connections | |
| Power (board power or TDP) | 300 W |
| GPU interconnect | PCIe Gen4 x16: 64 GB/s bidirectional; no NVLink |
| Multiple instances | Not stated by the manufacturer |
Note: Figures per GPU. The product page does not state compute performance; this comes from the datasheet (provisional). Max. 300 W. MIG: not supported; NVLink: not supported. Process node and launch date not published in the documents opened.
Alternatives to NVIDIA L40
Products with similar specifications from the same overview. This is not a ranking and not buying advice.
- AMD Radeon PRO W790048 GB · GDDR6 (ECC) · 0,86 TB/sCompare with NVIDIA L40 →
- NVIDIA L40S48 GB · GDDR6 (ECC) · 0,86 TB/sCompare with NVIDIA L40 →
- NVIDIA A3024 GB · HBM2 · 0,93 TB/sCompare with NVIDIA L40 →
- NVIDIA L424 GB · 0,3 TB/sCompare with NVIDIA L40 →
- NVIDIA A100 80GB PCIe80 GB · HBM2e · 1,94 TB/sCompare with NVIDIA L40 →
- AMD Instinct MI21064 GB · HBM2e · 1,6 TB/sCompare with NVIDIA L40 →
Who is the L40 meant for?
According to the manufacturer’s data, the L40 is meant for data centres, AI workloads and compute-heavy work. Whether it suits you depends on what you want to do with it. Use the rankings and comparisons on Clouwino to put it next to similar products. Clouwino points are derived from the specifications and are not a benchmark.
Key specifications of the L40
- Architecture: Ada Lovelace
- Form: PCIe, dual-slot (passief)
- FP32 compute: 90.5 TFLOPS
- FP16/BF16: 181.05 TFLOPS
- FP8: 362 TFLOPS
- Memory: 48 GB
- Memory type: GDDR6 (ECC)
- Memory bandwidth: 0.864 TB/s
All values come from the manufacturer’s official page. Where a value is missing, the manufacturer does not state it. Read how we compare.
Frequently asked questions
What is the fp32 compute of the L40?
According to the manufacturer, the fp32 compute of the NVIDIA L40 is 90.5 TFLOPS. Checked on 19 October 2026.
What is the fp16/bf16 of the L40?
According to the manufacturer, the fp16/bf16 of the NVIDIA L40 is 181.05 TFLOPS. Checked on 19 October 2026.
What is the fp8 of the L40?
According to the manufacturer, the fp8 of the NVIDIA L40 is 362 TFLOPS. Checked on 19 October 2026.
What are good alternatives to the L40?
The alternatives page for the L40 shows similar products based on the specifications. It is not a test and not a judgement of quality.
Last updated on
