3_daef569a-17bb-45b4-b8c9-f853466f298d

"Image may differ from actual product"

Roll over image to zoom in Click on image to zoom
3_daef569a-17bb-45b4-b8c9-f853466f298d

PNY NVL40STCGPU-KIT NVIDIA L40S 48GB GDDR6 PCIe Gen4 Passive GPU

Price:
$7,709.00 $11,500.00
PNYSKU: PSY7698064
Model Number: NVL40STCGPU-KIT
Condition: New
In stock:  Yes
Get Volume Pricing

Shop with confidence! Saitech Inc is an authorized PNY Partner Reseller.

     

PNY NVL40STCGPU-KIT NVIDIA L40S 48GB GDDR6 PCIe Gen4 Passive GPU

PNY NVL40STCGPU-KIT NVIDIA L40S 48GB GDDR6 PCIe Gen4 Passive GPU

SPECIAL ORDER ITEM

This product is available for special order and may take longer to ship.

AVAILABILITY & SHIPPING

We source this item directly from the OEM as a special order, which may result in longer lead times. We’ll keep you updated throughout the process.

NEED IT SOONER?

Contact us to explore expedited options or find a comparable in-stock alternative.

SPECIAL ORDER FULFILLMENT

Saitech Inc. sources special orders directly from manufacturers and authorized distributors, including hard-to-find, end-of-life, and build-to-order products. These items may have longer lead times, and our team will keep you informed throughout the process.

  • Overview

Overview

The NVIDIA L40S GPU is the definitive data center solution, purpose-built to deliver breakthrough performance across large language model (LLM) inference and training, computer graphics, rendering, and video workloads. Powered by the revolutionary NVIDIA Ada Lovelace architecture, the L40S sets a new benchmark for versatility, compute density, and energy-efficient performance in a passively cooled form factor optimized for high-density server deployments.

Specifications

Model Number:  NVL40STCGPU-KIT

Brand:  PNY

Alt Part Number(s):  NVL40STCGPU-KIT

Product Type:  Graphic Card 

Technical Description

Brand:  PNY Technologies

Model:  NVL40STCGPU-KIT

Chipset Manufacturer:  NVIDIA

Chipset:  L40S  

Number of CUDA Cores:  18176

Number of RT Cores:  142 

Number of Tensor Cores:  568 

Number of Display Ports:  4-Port

Standard Memory:  48 GB

Memory Technology:  GDDR6

Graphics Bus: PCIe Gen40x 16 

Power Connector:  16-pin

Max Power Consumption:  350W

Output Voltage:  208 - 230 V AC

Protocols:  Web, SNMP, Telnet

Designed for multi-modal acceleration, the NVIDIA L40S GPU excels across AI, graphics, and HPC pipelines. It supports Multi-Instance GPU (MIG) capabilities for maximum workload flexibility and resource optimization, allowing enterprises to partition GPU resources and run multiple tasks in parallel with secure isolation.

*

Product information, pricing, availability, images or discounts are subject to change without notice and may not always reflect the most current details. In the event of pricing or availability discrepancies, we reserve the right to cancel or adjust orders. Customers will be notified and may choose to proceed or cancel.

You may also like

Recently viewed

Frequently Asked Questions

Who should consider using the NVIDIA L40S GPU?

Professionals working in machine learning, data science, simulation, architectural rendering, and high-end graphics production should consider this GPU. Its compute and visualization capabilities help accelerate workflows, reduce processing bottlenecks, and manage complex datasets efficiently across technical and creative applications.

How does the L40S GPU contribute to overall system performance?

By offloading graphics and compute workloads from the CPU to specialized GPU cores, the L40S GPU frees up system resources, accelerates compute-intensive tasks, and enables faster data processing, which enhances overall responsiveness and throughput across demanding applications.

Why is the L40S GPU considered a high-end enterprise solution?

The NVIDIA L40S GPU is considered a high-end enterprise solution because it delivers exceptional graphics performance, compute acceleration, and AI capabilities in one package, making it suitable for mission-critical workloads, scalable deployments, and professional environments demanding dependable GPU power.

Is it suitable for cloud and data center deployments?

Yes, L40S GPU is suitable for cloud and data center deployments by offering scalable performance, remote access capability, and efficient resource sharing, helping organizations build robust infrastructure for visualization, AI services, and virtual computing.

How does it help reduce latency in interactive applications?

It reduces latency in interactive applications by accelerating frame rendering and compute tasks, ensuring rapid response times for simulation controls, virtual environments, and real-time feedback loops in design and analysis workflows.