Home NVIDIA GeForce RTX 4070 AD103

NVIDIA GeForce RTX 4070 AD103

This is a GPU manufactured with TSMC 5nm process, based on Nvidia Ada Lovelace architecture and released on Mar 2024. It features 45.9 billion transistors, 5888 CUDA cores and 12GB GDDR6X memory, with 36MB L2 cache, theoretical performance of 29.15TFLOPS, with total power consumption of 200W.

Graphics Card

[Report Issues]
Release Date
Mar 2024
Generation
GeForce 40
Type
Desktop
Bus Interface
PCIe 4.0 x16

Clock Speeds

[Report Issues]
Base Clock
1920MHz
Boost Clock
2475MHz
Memory Clock
1313MHz

Memory

[Report Issues]
Memory Size
12GB
Memory Type
GDDR6X
Memory Bus
192bit
Bandwidth
504.2 GB/s

Render Config

[Report Issues]
SM Count
46
Shading Units
5888
TMUs
184
ROPs
64
Tensor Cores
184
RT Cores
46
L1 Cache
128 KB (per SM)
L2 Cache
36MB

Theoretical Performance

[Report Issues]
Pixel Rate
158.4 GPixel/s
Texture Rate
455.4 GTexel/s
FP16 (half)
29.15 TFLOPS
FP32 (float)
29.15 TFLOPS
FP64 (double)
455.4 GFLOPS

Graphics Processor

[Report Issues]
GPU Name
AD103
GPU Variant
AD103-175-K1-A1
Architecture
Ada Lovelace
Foundry
TSMC
Process Size
5 nm
Transistors
45,900 million
Die Size
379mm²

Board Design

[Report Issues]
TDP
200W
Suggested PSU
550W
Outputs
1x HDMI 2.1
3x DisplayPort 1.4a
Power Connectors
1x 16-pin

Graphics Features

[Report Issues]
DirectX
12 Ultimate (12_2)
OpenGL
4.6
OpenCL
3.0
Vulkan
1.3
CUDA
8.9
Shader Model
6.7

Rankings

[Report Issues]
FP32 (float)
30.64 TFLOPS
NVIDIA L4 24 GB GDDR6
30.29 TFLOPS
29.77 TFLOPS
29.15 TFLOPS
29.15 TFLOPS
28.31 TFLOPS
NVIDIA RTX A5000 24 GB GDDR6
27.77 TFLOPS

Related Comparisons

© 2024 - TopCPU.net   Contact Us Privacy Policy