EXALIT

   +65 6591 8608

Friday18 August 2017

NVIDIA Tesla cards - P40 24GB

NVIDIA Tesla P40 accelerator is engineered to deliver the highest throughput for FP32 workload and Deep Learning servers.
It has 3840 CUDA cores with a peak FP32 throughput of 12 TFLOP/s.


It features additional instructions aimed at deep learning computations - INT8 vector dot products (IDP2A / IDP4A instructions), with a peak throughput of 47.0 INT8 TOP/s.

Tesla P40

 

 

Specification table:

Class Tesla GPGPU Accelerators
Model Tesla P40 24GB
NVIDIA Microarchitecture Pascal
CUDA cores 3840
Base Clock 1303 Mhz
GPU Boost Clock 1531 Mhz
INT8 TOP/s (Tera Operations Per Sec) 47.0
Single precision / FP32 12.0 TFlops
Double precision / FP64 0.35 TFlops
PCIe Gen 3.0
Memory size 24 GB GDDR5
Memory bandwidth 346 GB/s
Power 250W

For Nvidia Tesla P40 Datasheet please visit the following link.

JA Minisite

2857.orig.q75.o0 - Copy IntelTPP amd  tesla preferred partner  quadro partner  qctlogo e   Mellanox APAC Partner   1