Q0V80A - Refurbished - HPE NVIDIA Tesla P40 24GB Computational Accelerator

Condition: Refurbished

Requires 8P Keyed Cable Kit (871829-B21)

HPE NVIDIA Tesla P40 24GB Computational Accelerator (Part Number: Q0V80A) is a high-performance GPU designed for data center applications such as deep learning inference, machine learning, and high-performance computing (HPC). It leverages NVIDIA's Pascal architecture and is optimized for inference workloads.

Specifications

• GPU Architecture: NVIDIA Pascal (GP102)

• CUDA Cores: 3,840

• Base Clock: 1,300 MHz

• Boost Clock: 1,530 MHz

• Memory: 24 GB GDDR5

• Memory Interface: 384-bit

• Memory Bandwidth: 346 GB/s

• Interface: PCI Express 3.0 x16

• Form Factor: Dual-slot

• Cooling: Passive (requires adequate system airflow)

• Power Consumption: 250W (1x 8-pin PCIe power connector)

• ECC Support: Yes

• Display Outputs: None

Performance Highlights

• FP32 (Single Precision): 11.7 TFLOPS

• FP64 (Double Precision): 367 GFLOPS

• INT8 Inference Performance: 47 TOPS

Compatibility & Deployment

The Tesla P40 is compatible with systems that have a PCIe 3.0 x16 slot and sufficient airflow for passive cooling. It has been used in various server models, including:

• HPE Apollo pc40

• HPE ProLiant DL380 Gen9

• HPE ProLiant XL190r Gen9

• HPE Synergy 480 Gen10