NVIDIA L40S 48GB PCIe Ada Lovelace GPU
Brand: NVIDIA | Category: GPUs
SKU: 900-2G143-0000-000 | Part #: 900-2G143-0000-000 | MPN: 900-2G143-0000-000
Contact for Pricing — Request a Quote
About the NVIDIA L40S 48GB PCIe Ada Lovelace GPU
NVIDIA L40S 48GB PCIe Ada Lovelace GPU
The NVIDIA L40S is the highest-performance PCIe AI inference and graphics workload GPU in the Ada Lovelace generation. With 48GB GDDR6 ECC memory and 91.6 TFLOPS FP32 performance, the L40S fits in a standard PCIe 4.0 x16 slot (no SXM baseboard required), making it the most flexible accelerator for rack servers requiring AI inference, 3D rendering, virtual workstation (NVIDIA RTX Virtual Workstation), and video transcoding in a single form factor.
L40S vs H100 PCIe: Workload Split
The L40S excels at inference (especially when batching is limited), real-time graphics/rendering, and mixed AI+visualization workloads — while the H100 PCIe is optimised for training and large-batch inference. For UAE media, architecture, and engineering firms deploying GPU-accelerated remote workstations alongside inference APIs, the L40S's dual-purpose capability (rendering + inference) delivers more value than a training-only accelerator.
Key Specifications
| Parameter | Value |
|---|---|
| Architecture | Ada Lovelace (AD102) |
| CUDA Cores | 18,176 |
| Memory | 48GB GDDR6 ECC |
| Memory Bandwidth | 864 GB/s |
| FP32 Performance | 91.6 TFLOPS |
| INT8 (Tensor) | 733 TOPS |
| Interface | PCIe 4.0 x16 (passive, dual-slot) |
| TDP | 350W |
| NVLink | Not supported (PCIe form factor) |
| Display Output | 4× DisplayPort 1.4 |
Technical Specifications
| Brand | NVIDIA |
| Category | GPUs |
| SKU | 900-2G143-0000-000 |
| Part Number | 900-2G143-0000-000 |
| Condition | New |
| Capacity | 48GB |
| Interface | PCIe Gen4 x16 |
| Power Supply | 350W |
| GPU Support | NVIDIA L40S |
| AI Optimized | Yes |
| Architecture | Ada Lovelace |
| Max Memory | 48GB GDDR6 ECC |
| Memory Bandwidth | 864 GB/s |
| FP8 Performance | 733 TFLOPS |
| TDP | 350W |
| GPU Interface | PCIe Gen4 x16 |
| Use Case | AI + Graphics + Video |
Frequently Asked Questions about NVIDIA L40S 48GB PCIe Ada Lovelace GPU
What does the NVIDIA L40S 48GB PCIe Ada Lovelace GPU do?
The NVIDIA L40S 48GB PCIe Ada Lovelace GPU accelerates AI/ML training, inference, scientific HPC, and virtualization (vGPU) workloads. Typical deployments include LLM training clusters, computer-vision pipelines, financial risk modeling, and rendering farms.
What are the headline specs of the NVIDIA L40S 48GB PCIe Ada Lovelace GPU?
Key specifications for the NVIDIA L40S 48GB PCIe Ada Lovelace GPU: new condition; brand NVIDIA; architecture Ada Lovelace; model L40S; memory 48GB GDDR6 ECC; memory bandwidth 864 GB/s; fp8 performance 733 TFLOPS. Manufacturer part number 900-2G143-0000-000. For the full datasheet with electrical, environmental, and compliance details, contact our pre-sales engineering team.
Is the NVIDIA L40S 48GB PCIe Ada Lovelace GPU compatible with my infrastructure?
The NVIDIA L40S 48GB PCIe Ada Lovelace GPU requires a PCIe Gen4 or Gen5 x16 slot, server power adequate for the card's TDP, and CUDA/ROCm driver support in your hypervisor or bare-metal OS. Sales engineering will confirm chassis fit (1U/2U/4U), PCIe lane count, and PSU headroom before quoting.