GPU

NVIDIA Tesla P4

Edit@2 months ago

Intergrated Memory(VRAM)
Capacity

8 GB

(GDDR5 256-bit)

Bandwidth

192 GB/s

27 Token/s

Vector Compute
FP64
178.20 G
FP32
5.70 T
FP16
X
BF16
X
INT32
NI
INT8
X

NVIDIA Tesla P4 General-Purpose Float-Point performance (Vector Performance / Scalar Performance)

FP64: 178.20 GFLOPS

FP32: 5.70 TFLOPS

Matirx Compute
FP64
X
FP32
X
FP16
X
FP8
X
TF32
X
BF16
X
INT16
X
INT8
X
INT4
X

NVIDIA Tesla P4 AI performance (Tensor Performance / Matrix Performance)

Hardware Specs
NVIDIA Tesla P4 is a 16nm chip, has 7200 million transistors, launched by NVIDIA at 2016. It has 8 GB built-in(On-Board/On-Chip) memory with bandwidth up to 192 GB/s. It has 2560 general-purpose ALUs(CUDA cores/Shader cores).
Process Node
16 nm
Launch Year
2016
Vector(CUDA) Cores
2560
Matrix(Tensor) Cores
Core Frequency
886 ~ 1114 MHz
Cache
2MB