HPE870919-001

NVIDIATesla P40 24 GB PCIe GPU Accelerator

The HPE870919-001 features an NVIDIATesla P40 GPU with 24 GB of GDDR5 memory and a PCI Express 3.0 ×16 interface. Designed for AI inference,high-performance computing, and virtual workstation acceleration, it supports compute workloads across CUDA, OpenCL, and TensorRT. With 3,840 CUDA cores and up to ~12 TFLOPS of single-precision performance, this Acceleratorexcels at parallel processing and inference tasks at scale. The large memory capacity and memory bandwidth (~346 GB/s) make it well-suited for large neural network inference, graphics virtualization, and simulation workloads in HPEProLiant and Synergy server environments.

Features

Based on NVIDIAPascal architecture (GP102 GPU), the Tesla P40 delivers up to ~47 TOPS of INT8 inferencing throughput (ideal for quantized deep-learning models), along with ~11.7 TFLOPS FP32 compute performance. The board is passively cooled and rated at ~250 W TDP, requiring sufficient chassis airflow for thermal management. Equipped with 24 GB GDDR5 memory on a 384-bit bus, it sustains ~346 GB/s memory bandwidth for heavy workloads. The card supports hardware virtualization via NVIDIAGRID/vDWS profiles, up to 24 vGPU instances (1 GB profile) for remote desktops, plus GPU compute APIs like CUDA, OpenCL, and TensorRT. Compatible with HPEProLiant Gen9/Gen10 servers including DL380, DL360, Apollo, XL, and Synergy platforms.

Advanced Options

The Tesla P40 is optimized for enterprise-class AI inference and GPU virtualization workloads. Powered by passive cooling, it relies on server airflow (bidirectional supported) to maintain thermal stability. With high single-precision and INT8 throughput, it accelerates inference models drastically×for example, a single server with eight P40s can replace over 140 CPU-only servers in inference tasks.

Product Key Features

• NVIDIA Tesla P40 GPU Accelerator(Pascal architecture)
  

• 24 GB GDDR5 memory, 384-bit memory interface (~346 GB/s)
  

• 3,840 CUDA cores; ~11.7 TFLOPS FP32 compute throughput
  

• INT8 inference performance: ~47 TOPS
  

• Passive cooling (dual-slot PCIe board), 250 W max power
  

• Supports vGPU virtualization (up to 24 users), CUDA, OpenCL, TensorRT

---

Reliability & Performance

Support for NVIDIAvGPU enables remote workstation deployments and multi-user GPU sharing, enhancing data center efficiency. ECC memory protection and compliance with HPE-certified server platforms ensure operational reliability. Its deployment in HPESynergy and ProLiant systems facilitates streamlined installation, compatibility, and support for long-term use.

Key Features :

• NVIDIATesla P40 GPU with Pascal GP102 architecture

• 24 GB GDDR5 memory, 384-bit interface (~346 GB/s)

• Compute performance: ~11.7 TFLOPS FP32, ~47 TOPS INT8 inference

• Passive cooling; dual-slot, PCIe 3.0 ×16 form factor

• Supports NVIDIAvGPU (up to 24 virtual GPUs), CUDA, OpenCL, TensorRT

• Certified for HPEProLiant Gen9/Gen10, Synergy, Apollo and XL platforms

Specification	Details
Model Number	870919-001 (Tesla P40, HP-branded)
GPU Architecture	NVIDIAPascal (GP102) Accelerator
CUDA Cores	3,840
Memory	24 GB GDDR5
Memory Interface	384-bit; ~346 GB/s bandwidth
Compute Throughput	~11.7 TFLOPS FP32 / ~47 TOPS INT8
System Interface	PCI Express 3.0 ×16
Power Consumption	~250 W TDP, passive-cooled
Cooling	Passive, requires system airflow
Dimensions	~1.4″ × 4.4″ × 10.5″ (dual-slot)
Weight	~2.35 lb (~1 kg)
Virtualization Support	NVIDIAvGPU (up to 24 instances, 1 GB profile)
Compute APIs	CUDA, OpenCL, TensorRT
Compatibility	HPEProLiant Gen9/Gen10, Synergy 480 Gen10, Apollo, XL series
Certifications	ECC memory, HPEcertified
Warranty	~90-day refurb; varies by reseller