GPU Wiki

GeForce RTX 4070 Ti vs GeForce RTX 5070 Ti

Swap Clear comparison
Edit GPUs
Add a third or fourth GPU

Key Differences

  • No game was measured on every one of these cards on one shared test system, so there is no speed comparison to make.
  • RTX 5070 Ti has the most memory: 16 GB against 12 GB.
  • RTX 5070 Ti has the most memory bandwidth: 896 GB/s against 504.2 GB/s.
  • RTX 4070 Ti draws the least power: 285 W against 300 W.
Spec GeForce RTX 4070 Ti GeForce RTX 5070 Ti
Overview
Product name NVIDIA GeForce RTX 4070 Ti NVIDIA GeForce RTX 5070 Ti
Architecture Ada Lovelace Blackwell 2.0
Generation / family GeForce RTX 40 Series GeForce RTX 50 Series
GPU codename AD104 GB203
Launch date 2023-01-05 2025-02-20
Market segment Desktop, High End in GeForce RTX 40 Series Desktop, High End in GeForce RTX 50 Series
Board type / entry type Reference GPU Reference GPU
GPU
Process node TSMC 5 nm TSMC 5 nm
Die size 294 mm² 378 mm²
Transistor count 35.8 billion 45.6 billion
GPU chip AD104 GB203
SM / multiprocessor count 60 70
CUDA Cores 7680 8960
Tensor Cores 240 280
RT Cores 60 70
ROPs 80 96
TMUs 240 280
Clocks
Base / Core clock 2,310 MHz 2,295 MHz
Boost clock 2,610 MHz 2,452 MHz
Memory clock 1,313 MHz 1,750 MHz
Memory
Memory size 12 GB 16 GB
Memory type GDDR6X GDDR7
Memory bus width 192-bit 256-bit
Memory bandwidth 504.2 GB/s 896 GB/s
Bus / Interface
Bus interface PCIe 4.0 x16 PCIe 5.0 x16
AGP / PCI / PCIe distinction PCIe PCIe
Power
TDP / TBP / board power 285 W 300 W
Recommended PSU 700 W 750 W
Power connectors 1x PCIe Gen 5 16-pin 1x PCIe Gen 5 16-pin
Slot width Dual-slot, 285 mm length, 42 mm width Dual-slot, 304 mm length, 48 mm width
Display / Media
Display outputs 1x HDMI 2.1, 3x DisplayPort 1.4a 1x HDMI 2.1b, 3x DisplayPort 2.1b
Features
DirectX support DirectX 12.2 DirectX 12.2
OpenGL support OpenGL 4.6 OpenGL 4.6
Vulkan support Vulkan 1.4 Vulkan 1.4
CUDA support 8.9 12.0
Shader Model 6.8 6.8
Ray tracing support Hardware RT units present Hardware RT units present
Performance / Benchmarks
FP32 throughput 40.09 TFLOPS 43.94 TFLOPS
FP16 throughput 40.09 TFLOPS 43.94 TFLOPS
Muted cells indicate unavailable, not applicable, or not directly comparable specifications.

Frame rates compare runs of the same game at the same quality preset on one test system. DLAA counts as native resolution, because it applies anti-aliasing at full render resolution rather than upscaling; DLSS and FSR upscaling are never counted as native. Where a capture named ray tracing without a readable on or off value, it is read as on.

Measured game results

Gaming benchmarks

Reviewed FPS from games these cards were measured on under one shared test system and resolution. Every game is compared against itself, so each result is a like-for-like pair. These results are separate from theoretical compute specifications.

No shared approved gaming benchmark yet

The selected GPUs do not currently share reviewed measurements from the same game, test system and resolution. No result was estimated or filled in.