NVIDIA L40 48GB
Ada Lovelace PCIe 2023 4nm
FP32
90.5
TFLOPS
VRAM
48 GB
GDDR6
TDP
300 W
301.7 TFLOPS/kW
Bandwidth
864 GB/s
memory
Performance Metrics
Peak theoretical throughput by precision type
| Precision | Description | Bits | Peak TFLOPS | Efficiency |
|---|---|---|---|---|
| INT8 | 8-bit integer | 8 | 362.0 | 1.207 TFLOPS/W |
| INT4 | 4-bit integer | 4 | 724.0 | 2.413 TFLOPS/W |
| FP8 | 8-bit floating point | 8 | 362.0 | 1.207 TFLOPS/W |
| FP16 | 16-bit floating point | 16 | 181.1 | 0.604 TFLOPS/W |
| BF16 | Brain Float 16 | 16 | 181.1 | 0.604 TFLOPS/W |
| TF32 | TensorFloat-32 | 32 | 90.5 | 0.302 TFLOPS/W |
| FP32 | 32-bit floating point | 32 | 90.5 | 0.302 TFLOPS/W |
FP8 Efficiency
1.207 TFLOPS/W
362.0 TFLOPS / 300W
FP16 Efficiency
0.604 TFLOPS/W
181.1 TFLOPS / 300W
FP32 Efficiency
0.302 TFLOPS/W
90.5 TFLOPS / 300W
Power Specifications
TDP
300 W
Max Power
345 W
Power Connector
PCIe 8-pin
Cooling
Air
Memory Specifications
Capacity
48 GB
Type
GDDR6
Bandwidth
864 GB/s
Interface
--
Hardware & Design
Form Factor
PCIe
Architecture
Ada Lovelace
Process Node
4nm
Launch Year
2023
Variant
48GB
Market Segment
Data Center
Common Use Cases
AI Inference Video Processing Edge Computing Computer Vision
The NVIDIA L40 is optimized for high-performance computing tasks with Ada Lovelace architecture delivering 90.5 TFLOPS of compute power.
Where to Rent
Compare cloud providers offering on-demand GPU instances for AI training, inference, and HPC workloads.
Browse GPU Cloud ProvidersSimilar GPUs
Stay Updated on GPU Releases
Get notified when new GPUs are added or specifications are updated.
Loading verification...
No spam, unsubscribe anytime.