Report an Error

NVIDIA GP100

Default
GP100
GP100-890-A1
GP100-890-A1
GP100-893-A1
GP100-893-A1
Die Shot
Fritzchens Fritz
Die Shot
Block Diagram
Block Diagram
SM Diagram
SM Diagram
NVIDIA's GP100 GPU uses the Pascal architecture and is made using a 16 nm production process at TSMC. With a die size of 610 mm² and a transistor count of 15,300 million it is a very big chip. GP100 supports DirectX 12 (Feature Level 12_1). It features 3584 shading units, 224 texture mapping units and 96 ROPs.
Further reading: GP100 Architecture Whitepaper

Graphics Processor

GPU Name
GP100
Codename
NV130
Architecture
Pascal
Foundry
TSMC
Process Size
16 nm
Transistors
15,300 million
Density
25.1M / mm²
Die Size
610 mm²
Released
Apr 5th, 2016

Graphics Features

DirectX
12 (12_1)
OpenGL
4.6
OpenCL
1.2
Vulkan
1.2.131
CUDA
6.0
Shader Model
6.4
PureVideo HD
VP8
VDPAU
Feature Set H

Render Config

Shading Units
3584
TMUs
224
ROPs
96
SM Count
56
SFUs
960
TPCs
30
GPCs
6
L1 Cache
24 KB per SM
L2 Cache
4096 KB
Max. TDP
300 W

All Pascal GPUs

NVIDIA GPU Architecture History

Graphics cards using the NVIDIA GP100 GPU

Name Chip Memory Shaders TMUs ROPs Base Clock Boost Clock Memory Clock
NVIDIA Tesla P100 PCIe 16 GB GP100-893-A1 16 GB 3584 224 96 1190 MHz 1329 MHz 715 MHz
NVIDIA Tesla P100 PCIe 12 GB 12 GB 3584 224 96 1190 MHz 1329 MHz 715 MHz
NVIDIA Quadro GP100 16 GB 3584 224 96 1304 MHz 1442 MHz 715 MHz
NVIDIA Tesla P100 SXM2 GP100-890-A1 16 GB 3584 224 96 1328 MHz 1480 MHz 704 MHz
NVIDIA Tesla P100 DGXS 16 GB 3584 224 96 1328 MHz 1480 MHz 715 MHz