GPUDatacenterReleased —

NVIDIA L40 CNX

384-bit · 18,176 cores

The L40 CNX is a data center variant of the NVIDIA L40, based on the full AD102 Ada Lovelace GPU with 18,176 CUDA cores, 568 fourth-generation Tensor Cores and 142 third-generation RT Cores. It is aimed at servers running AI inference, graphics and virtualization workloads.

General

Manufacturer: NVIDIA
Model: L40 CNX
Segment: Datacenter
GPU chip: AD102
Architecture: Ada Lovelace
Release date: —
Price: Search on Amazon

Memory

Memory size: —
Memory type: GDDR6
Bus width: 384-bit
Bandwidth: 864 GB/s
Memory clock: 2,250 MHz

Clocks & silicon

GPU clock: —
Boost clock: —
Shader units: 18,176 CUDA Cores
Tensor / AI cores: 568
RT cores: 142
Transistors: 76.3 billion
Technology (nm): 5 nm
TSMC 4N (5 nm-class)
TDP / TBP: —

Interface & APIs

Bus interface: —
SLI / multi-GPU: —
DirectX: 12 Ultimate (12_2)
OpenGL: 4.6
Vulkan: 1.3
Other APIs: CUDA 8.9, OpenCL 3.0

AI support

Precisions: FP8FP16BF16INT8INT4TF32FP64
AI accelerator: 4th Gen Tensor Cores (568) + 3rd Gen RT Cores (142)
Peak AI TOPS: —
Notes: Fourth-generation Ada Lovelace Tensor Cores support FP8, FP16, BF16, TF32, INT8 and INT4 with structured sparsity. FP64 runs at 1/64 rate.