Are you looking for a data center-level GPU to power AI inference, rendering, graphics, digital twins, and HPC workloads? Then, explore the advanced capabilities of the NVIDIA L40 GPU 48GB, an enterprise-grade accelerator built on the Ada Lovelace architecture. This powerhouse boasts 48 GB of GDDR6 ECC memory, 18,176 CUDA cores, 568 4th-generation Tensor Cores, and 142 3rd-generation RT Cores. With 864 GB/S of memory bandwidth and a PCIe Gen4 x16 interface, it delivers high performance for workloads on servers from Dell, HPE, Lenovo, Supermicro, and Cisco. ServerBasket offers competitive NVIDIA L40 GPU prices, fast shipping, technical support, and a 1-year warranty on the NVIDIA L40 GPU.

Technical Specifications

GPU Architecture	NVIDIA Ada Lovelace Architecture
GPU Memory	48GB GDDR6 with ECC
CUDA Cores	18176
Memory Bandwidth	864 GB/s
Interconnect Interface	PCIe Gen4 x16; 64 GB/s Bi-Directional
RT Cores	142 (3rd Generation)
Tensor Cores	568 (4th Generation)
RT Core Performance	209 TFLOPS
FP32 Performance	90.5 TFLOPS
TF32 Tensor Core Performance	90.5 \| 181 TFLOPS
BFLOAT16 Tensor Core Performance	181.05 \| 362.1 TFLOPS
FP16 Tensor Core Performance	181.05 \| 362.1 TFLOPS
FP8 Tensor Core Performance	362 \| 724 TFLOPS
Peak INT8 Tensor TOPS	362 \| 724 TOPS
Peak INT4 Tensor TOPS	724 \| 1448 TOPS
Form Factor	4.4" (H) × 10.5" (L), Dual Slot
Display Ports	4 × DisplayPort 1.4a
Max Power Consumption	300W
Power Connector	16-pin
Thermal Solution	Passive
Virtual GPU (vGPU) Support	Yes
vGPU Profiles Supported	See Virtual GPU Licensing Guide
NVENC / NVDEC	3 × NVENC, 3 × NVDEC (Includes AV1 Encode & Decode)
Secure Boot with Root of Trust	Yes
NEBS Ready	Level 3