Home NVIDIA GeForce RTX 5070 SUPER

NVIDIA GeForce RTX 5070 SUPER

This is a GPU manufactured with TSMC 5nm process, based on Nvidia Blackwell 2.0 architecture and released on 2026. It features 31.1 billion transistors, 6400 CUDA cores and 18GB GDDR7 memory, with 48MB L2 cache, theoretical performance of 32.15TFLOPS, with total power consumption of 275W.

Graphics Card

[Report Issues]
Release Date
Jan 2026
Generation
GeForce 50
Type
Desktop
Bus Interface
PCIe 5.0 x16

Clock Speeds

[Report Issues]
Base Clock
2325 MHz
Boost Clock
2512 MHz
Memory Clock
1750 MHz

Memory

[Report Issues]
Memory Size
18GB
Memory Type
GDDR7
Memory Bus
192bit
Bandwidth
672.0GB/s

Render Config

[Report Issues]
SM Count
50
Shading Units
6400
TMUs
200
ROPs
80
Tensor Cores
200
RT Cores
50
L1 Cache
128 KB (per SM)
L2 Cache
48 MB

Theoretical Performance

[Report Issues]
Pixel Rate
201.0 GPixel/s
Texture Rate
502.4 GTexel/s
FP16 (half)
32.15 TFLOPS
FP32 (float)
32.15 TFLOPS
FP64 (double)
502.4 GFLOPS

Graphics Processor

[Report Issues]
GPU Name
GB205
GPU Variant
GB205-400-A1
Architecture
Blackwell 2.0
Foundry
TSMC
Process Size
5 nm
Transistors
31.1 billion
Die Size
263 mm²

Board Design

[Report Issues]
TDP
275W
Suggested PSU
600 W
Outputs
1x HDMI 2.1b
3x DisplayPort 2.1b
Power Connectors
1x 16-pin

Graphics Features

[Report Issues]
DirectX
12 Ultimate (12_2)
OpenGL
4.6
OpenCL
3.0
Vulkan
1.4
CUDA
12.0
Shader Model
6.8

Rankings

[Report Issues]
FP32 (float)
NVIDIA RTX 5000 Max-Q Ada Generation
32.69 TFLOPS
NVIDIA GeForce RTX 5070 SUPER
32.15 TFLOPS
AMD Radeon PRO W7700
31.95 TFLOPS
NVIDIA GeForce RTX 5090 Mobile
31.8 TFLOPS
NVIDIA A10G
NVIDIA A10G 12 GB GDDR6
31.52 TFLOPS
© 2026 - TopCPU.net