Q2N68A - Refurbished - HPE NVIDIA Tesla V100 PCIe 16GB Computational Accelerator

Condition: Refurbished

Requires 8P Keyed Cable Kit (871829-B21). Need just the cable? The 869820-001 cable from that kit is available separately for $49.

The HPE NVIDIA Tesla V100 PCIe 16GB Computational Accelerator (Part Number: Q2N68A) is a high-performance GPU designed for data center applications such as AI training, high-performance computing (HPC), and deep learning. It leverages NVIDIA's Volta architecture and High Bandwidth Memory (HBM2) to deliver exceptional computational capabilities.

Specifications

• GPU Architecture: NVIDIA Volta

• CUDA Cores: 5,120

• Tensor Cores: 640

• Base Clock: 1,245 MHz

• Boost Clock: 1,380 MHz

• Memory: 16 GB HBM2

• Memory Interface: 4,096-bit

• Memory Bandwidth: 900 GB/s

• Interface: PCI Express 3.0 x16

• Form Factor: Dual-slot, full-height (267 mm x 112 mm x 38 mm)

• Cooling: Passive (requires adequate system airflow)

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

• ECC Support: Yes (enabled by default)

• Display Outputs: None

Performance Highlights

• FP64 (Double Precision): 7.0 TFLOPS

• FP32 (Single Precision): 14.0 TFLOPS

• Tensor Performance: 112 TFLOPS

Compatibility & Deployment

The Tesla V100 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 ProLiant DL380 Gen10

• HPE ProLiant XL190r Gen10

• HPE ProLiant XL270d Gen9