Q0V80A - Refurbished - HPE NVIDIA Tesla P40 24GB Computational Accelerator
Warranty: Lifetime
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.
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
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.
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