AI hardware
NVIDIA L40
FP16 throughput
1.8×10¹⁴ FLOP/s
Memory
44.7 GB
TDP
300 W
NVIDIA L40 is an AI accelerator designed by NVIDIA (GPU), released in October 2022.
It delivers 1.8×10¹⁴ FLOP/s of tensor FP16/BF16 throughput. It carries 44.7 GB of memory at 805 GB/s of bandwidth. Its thermal design power is 300 W.
Full record
- Manufacturer
- NVIDIA
- Type
- GPU
- Release date
- Oct 13, 2022
- Tensor FP16/BF16
- 1.8×10¹⁴ FLOP/s
- FP32
- 9.1×10¹³ FLOP/s
- FP64
- 1.4×10¹² FLOP/s
- INT8
- 3.6×10¹⁴ OP/s
- Memory
- 44.7 GB
- Memory bandwidth
- 805 GB
- TDP
- 300 W
- Transistors (millions)
- 76,300
- Die size (mm2)
- 608
- Tensor cores
- 568
More from NVIDIA
SourceEpoch AI, 'Machine Learning Hardware'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/machine-learning-hardware. Licensed under CC BY 4.0.
← All ai hardware