Skip to product information
1 of 1

Cisco

Cisco UCSX-GPU-A40 | A40 Data Center GPU, 48GB GDDR6 ECC, PCIe Gen4, 300W Passive

SKU:UCSX-GPU-A40

Stock Status: Enquire

Request Quote
Sale Sold out
Taxes included. Shipping calculated at checkout.

Description

The NVIDIA A40 is a powerful data center GPU built on Ampere architecture, designed for AI inference, rendering, virtualization, and high-performance computing workloads. With 48GB GDDR6 ECC memory, 10,752 CUDA cores, and PCIe Gen4 support, it delivers exceptional performance for graphics-intensive applications, real-time ray tracing, and AI acceleration. The passively cooled design integrates seamlessly into Cisco UCS X-Series modular systems for enterprise-scale deployments.

Features

NVIDIA Ampere architecture with 10,752 CUDA cores for parallel processing
- 48GB GDDR6 ECC memory for data integrity and reliability in enterprise applications
- 3rd generation Tensor Cores for accelerated AI and deep learning inference
- 2nd generation RT Cores for real-time ray tracing with up to 2x throughput vs. previous generation
- PCIe Gen4 x16 interface for high-speed data transfers (backward compatible with Gen3)
- NVLink support for multi-GPU configurations with up to 96GB combined memory
- NVIDIA vGPU software support for virtual workstations and multi-user environments
- SR-IOV capability for hardware-based GPU virtualization
- Passive cooling design optimized for data center airflow
- ECC memory protection for mission-critical compute workloads
- DisplayPort 1.4a support with 8K resolution capability and Display Stream Compression
- NVIDIA GPUDirect RDMA and GPUDirect Storage for optimized data movement
- Configurable Max-Q operation mode for power efficiency optimization
- Full-height, dual-slot form factor for dense server deployments
- Includes 8-pin PCIe power cable for simplified installation

Warranty

All products sold by XS Network Tech include a 12-month warranty on both new and used items. Our in-house technical team thoroughly tests used hardware prior to sale to ensure enterprise-grade reliability.

All technical data should be verified on the manufacturer data sheets.

View full details

specs-tabs

Collapsible content

Technical Specifications

FAQs

Technical Specifications

GPU: NVIDIA A40 (Ampere architecture)
- Memory: 48GB GDDR6 with ECC
- Memory Interface: 384-bit
- CUDA Cores: 10,752
- Tensor Cores: 336 (3rd generation)
- RT Cores: 84 (2nd generation)
- Peak FP32 Performance: 37.4 TFLOPS
- Interface: PCIe Gen4 x16 (backward compatible with PCIe Gen3)
- Power Consumption: 300W (TDP)
- Cooling: Passive heatsink
- Form Factor: Full-height, dual-slot
- Power Connector: 8-pin PCIe auxiliary power (cable included)
- NVLink Support: Yes (NVLink bridge sold separately, enables up to 96GB combined memory with two A40 GPUs)
- Display Outputs: 3x DisplayPort 1.4a (disabled by default for virtualization; can be enabled via software)
- Virtualization: Supports NVIDIA vGPU and SR-IOV
- Dimensions: 4.4 in (H) x 10.5 in (L), dual-slot width
- Certifications: RoHS compliant

FAQs

Q: What systems is the UCSX-GPU-A40 compatible with?
A: This GPU is designed for Cisco UCS X-Series modular systems, specifically the UCS X440p PCIe Node paired with UCS X210c M6, M7, or X410c M7 compute nodes. It requires both UCS 9416 X-Fabric modules for PCIe Gen4 connectivity.

Q: What types of workloads is the A40 optimized for?
A: The A40 excels at AI inference, machine learning training, real-time ray tracing, 3D rendering, virtual workstations, CAD/CAE simulation, digital twins, virtual production, and high-performance computing tasks. Its 48GB ECC memory handles large datasets efficiently.

Q: Can I connect multiple A40 GPUs together?
A: Yes, you can connect two A40 GPUs using the NVIDIA NVLink bridge (sold separately) to scale memory to 96GB and achieve GPU-to-GPU data transfer rates up to 112.5 GB/s total bandwidth, ideal for larger workloads.

Q: Does the A40 support display outputs?
A: The A40 has three DisplayPort 1.4a outputs, but they are disabled by default for virtualization mode (required for NVIDIA vGPU and SR-IOV). Display outputs can be enabled via NVIDIA management software tools if needed.

Q: What is the difference between passive and active cooling?
A: This A40 uses passive cooling (heatsink only, no fans), designed for data center servers with chassis-level airflow management. Passive designs reduce mechanical failure points and noise while relying on the server's cooling infrastructure.

Recently Viewed

  • Request a Quote

    Looking for competitive pricing? Submit a request, and our team will provide a tailored quote that fits your needs.

  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Call us Now  
  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Contact us