Architecture
NVIDIA Ampere 8nm
Most Powerful Variant
GA102-300
Cores : TMUs : ROPs
10496 : 328 : 112
Typical Memory Size
24576 MB
Typical Memory Bus
384 -bit
GA102 GPU
Overview
Manufacturer
NVIDIA
Release Date
September 24th, 2020
Graphics Processing Unit
Architecture
Ampere
Fabrication Process
8 nm
Die Size
628 mm2
Transistors Count
28.3B
Tensor Cores
328
RT Cores
82
CUDAs
10496
TMUs
328
ROPs
112
Clocks
Base Clock
1395 MHz
Memory Configuration
Memory Size
24576 MB
Memory Type
GDDR6X
Memory Bus Width
384-bit
Memory Bandwidth
935.8 GB/s
Performance
Pixel Fillrate
189.8 GPixels/s
Texture Fillrate
556 GTexel/s
SPFP Performance
35.6 TFLOPS
Performance per W
101.7 GFLOPS/W
Performance per mm2
56.7 GFLOPS/mm2
Physical
Interface
PCI-Express 4.0 x16
Typical Thermal Design Power
350 W
Maximum Temperature
93 °C
SLI Support
2-way
API Support
DirectX
12.2
Vulkan
1.2
OpenGL
4.6
OpenCL
2.1
| Model | GPU Model | Unified Cores | Boost Clock | Memory Clock | Memory Config. | |
|---|---|---|---|---|---|---|
 | NVIDIA A40 | GA102 | TBC | TBC | TBC | TBC |
 | NVIDIA RTX A6000 | GA102 | TBC | TBC | TBC | TBC |
 | NVIDIA GeForce RTX 3090 | GA102-300 | ||||
 | NVIDIA GeForce RTX 3080 20GB | GA102-200 | ||||
 | NVIDIA GeForce RTX 3080 | GA102-200 | ||||
 | NVIDIA TITAN RTX AMPERE | GA102-400 | TBC | TBC | TBC | TBC |