GPU Wiki
GeForce RTX 4070 Ti SUPER vs GeForce RTX 4070 SUPER
Edit GPUs
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 4070 Ti SUPER has the most memory: 16 GB against 12 GB.
- RTX 4070 Ti SUPER has the most memory bandwidth: 672.3 GB/s against 504.2 GB/s.
- RTX 4070 SUPER draws the least power: 220 W against 285 W.
| Spec | GeForce RTX 4070 Ti SUPER | GeForce RTX 4070 SUPER |
|---|---|---|
| Overview | ||
| Product name | NVIDIA GeForce RTX 4070 Ti SUPER | NVIDIA GeForce RTX 4070 SUPER |
| Architecture | Ada Lovelace | Ada Lovelace |
| Generation / family | GeForce RTX 40 Series | GeForce RTX 40 Series |
| GPU codename | AD103 | AD104 |
| Launch date | 2024-01-24 | 2024-01-17 |
| Market segment | Desktop, High End in GeForce RTX 40 Series | Desktop in GeForce RTX 40 Series |
| Board type / entry type | Reference GPU | Reference GPU |
| GPU | ||
| Process node | TSMC 5 nm | TSMC 5 nm |
| Die size | 379 mm² | 294 mm² |
| Transistor count | 45.9 billion | 35.8 billion |
| GPU chip | AD103 | AD104 |
| SM / multiprocessor count | 66 | 56 |
| CUDA Cores | 8448 | 7168 |
| Tensor Cores | 264 | 224 |
| RT Cores | 66 | 56 |
| ROPs | 96 | 80 |
| TMUs | 264 | 224 |
| Clocks | ||
| Base / Core clock | 2,340 MHz | 1,980 MHz |
| Boost clock | 2,610 MHz | 2,475 MHz |
| Memory clock | 1,313 MHz | 1,313 MHz |
| Memory | ||
| Memory size | 16 GB | 12 GB |
| Memory type | GDDR6X | GDDR6X |
| Memory bus width | 256-bit | 192-bit |
| Memory bandwidth | 672.3 GB/s | 504.2 GB/s |
| Bus / Interface | ||
| Bus interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| AGP / PCI / PCIe distinction | PCIe | PCIe |
| Power | ||
| TDP / TBP / board power | 285 W | 220 W |
| Recommended PSU | 700 W | 650 W |
| Power connectors | 1x PCIe Gen 5 16-pin | 1x PCIe 8-pin or 1x PCIe Gen 5 16-pin |
| Slot width | Triple-slot, 310 mm length, 61 mm width | Dual-slot, 267 mm length, 42 mm width |
| Display / Media | ||
| Display outputs | 1x HDMI 2.1, 3x DisplayPort 1.4a | 1x HDMI 2.1, 3x DisplayPort 1.4a |
| 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 | 8.9 |
| Shader Model | 6.8 | 6.8 |
| Ray tracing support | Hardware RT units present | Hardware RT units present |
| Performance / Benchmarks | ||
| FP32 throughput | 44.1 TFLOPS | 35.48 TFLOPS |
| FP16 throughput | 44.1 TFLOPS | 35.48 TFLOPS |
Overview
Product name
GeForce RTX 4070 Ti SUPER
NVIDIA GeForce RTX 4070 Ti SUPER
GeForce RTX 4070 SUPER
NVIDIA GeForce RTX 4070 SUPER
Architecture
GeForce RTX 4070 Ti SUPER
Ada Lovelace
GeForce RTX 4070 SUPER
Ada Lovelace
Generation / family
GeForce RTX 4070 Ti SUPER
GeForce RTX 40 Series
GeForce RTX 4070 SUPER
GeForce RTX 40 Series
GPU codename
GeForce RTX 4070 Ti SUPER
AD103
GeForce RTX 4070 SUPER
AD104
Launch date
GeForce RTX 4070 Ti SUPER
2024-01-24
GeForce RTX 4070 SUPER
2024-01-17
Market segment
GeForce RTX 4070 Ti SUPER
Desktop, High End in GeForce RTX 40 Series
GeForce RTX 4070 SUPER
Desktop in GeForce RTX 40 Series
Board type / entry type
GeForce RTX 4070 Ti SUPER
Reference GPU
GeForce RTX 4070 SUPER
Reference GPU
GPU
Process node
GeForce RTX 4070 Ti SUPER
TSMC 5 nm
GeForce RTX 4070 SUPER
TSMC 5 nm
Die size
GeForce RTX 4070 Ti SUPER
379 mm²
GeForce RTX 4070 SUPER
294 mm²
Transistor count
GeForce RTX 4070 Ti SUPER
45.9 billion
GeForce RTX 4070 SUPER
35.8 billion
GPU chip
GeForce RTX 4070 Ti SUPER
AD103
GeForce RTX 4070 SUPER
AD104
SM / multiprocessor count
GeForce RTX 4070 Ti SUPER
66
GeForce RTX 4070 SUPER
56
CUDA Cores
GeForce RTX 4070 Ti SUPER
8448
GeForce RTX 4070 SUPER
7168
Tensor Cores
GeForce RTX 4070 Ti SUPER
264
GeForce RTX 4070 SUPER
224
RT Cores
GeForce RTX 4070 Ti SUPER
66
GeForce RTX 4070 SUPER
56
ROPs
GeForce RTX 4070 Ti SUPER
96
GeForce RTX 4070 SUPER
80
TMUs
GeForce RTX 4070 Ti SUPER
264
GeForce RTX 4070 SUPER
224
Clocks
Base / Core clock
GeForce RTX 4070 Ti SUPER
2,340 MHz
GeForce RTX 4070 SUPER
1,980 MHz
Boost clock
GeForce RTX 4070 Ti SUPER
2,610 MHz
GeForce RTX 4070 SUPER
2,475 MHz
Memory clock
GeForce RTX 4070 Ti SUPER
1,313 MHz
GeForce RTX 4070 SUPER
1,313 MHz
Memory
Memory size
GeForce RTX 4070 Ti SUPER
16 GB
GeForce RTX 4070 SUPER
12 GB
Memory type
GeForce RTX 4070 Ti SUPER
GDDR6X
GeForce RTX 4070 SUPER
GDDR6X
Memory bus width
GeForce RTX 4070 Ti SUPER
256-bit
GeForce RTX 4070 SUPER
192-bit
Memory bandwidth
Bus / Interface
Bus interface
GeForce RTX 4070 Ti SUPER
PCIe 4.0 x16
GeForce RTX 4070 SUPER
PCIe 4.0 x16
AGP / PCI / PCIe distinction
GeForce RTX 4070 Ti SUPER
PCIe
GeForce RTX 4070 SUPER
PCIe
Power
TDP / TBP / board power
GeForce RTX 4070 Ti SUPER
285 W
GeForce RTX 4070 SUPER
220 W
Recommended PSU
GeForce RTX 4070 Ti SUPER
700 W
GeForce RTX 4070 SUPER
650 W
Power connectors
GeForce RTX 4070 Ti SUPER
1x PCIe Gen 5 16-pin
GeForce RTX 4070 SUPER
1x PCIe 8-pin or 1x PCIe Gen 5 16-pin
Slot width
GeForce RTX 4070 Ti SUPER
Triple-slot, 310 mm length, 61 mm width
GeForce RTX 4070 SUPER
Dual-slot, 267 mm length, 42 mm width
Display / Media
Display outputs
GeForce RTX 4070 Ti SUPER
1x HDMI 2.1, 3x DisplayPort 1.4a
GeForce RTX 4070 SUPER
1x HDMI 2.1, 3x DisplayPort 1.4a
Features
DirectX support
GeForce RTX 4070 Ti SUPER
DirectX 12.2
GeForce RTX 4070 SUPER
DirectX 12.2
OpenGL support
GeForce RTX 4070 Ti SUPER
OpenGL 4.6
GeForce RTX 4070 SUPER
OpenGL 4.6
Vulkan support
GeForce RTX 4070 Ti SUPER
Vulkan 1.4
GeForce RTX 4070 SUPER
Vulkan 1.4
CUDA support
GeForce RTX 4070 Ti SUPER
8.9
GeForce RTX 4070 SUPER
8.9
Shader Model
GeForce RTX 4070 Ti SUPER
6.8
GeForce RTX 4070 SUPER
6.8
Ray tracing support
GeForce RTX 4070 Ti SUPER
Hardware RT units present
GeForce RTX 4070 SUPER
Hardware RT units present
Performance / Benchmarks
FP32 throughput
FP16 throughput
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.
