Skip to content

NVIDIA · Ada Lovelace · 16 GB

RTX 4070 Ti SUPER

16 GB GDDR6X · 256-bit · 8448 shaders · 285 W

1080p
65%
1440p
58%
4K
49%
Amazon
$1389
$1389
Unavailable Updated 1 week, 3 days ago
Gaming 1080p64.6%
Gaming 1440p57.8%
Gaming 4K48.8%
1440p value41.2

Written analysis

RTX 4070 Ti SUPER at a glance

The RTX 4070 Ti SUPER is based on the Ada Lovelace architecture and is manufactured using the 5 nm process at TSMC. It ships with 16 GB of GDDR6X on a 256-bit bus , about 672 GB/s. Shader count is 8448 across 56 CU/SM. The board interface is PCIe 4.0 x16.

Reference board power is 285 W with 1x 16-pin. This is manageable in a ventilated mid-tower, while compact builds require closer clearance and thermal checks.

VRAM capacity is strong at 16 GB. Comfortable for 1440p ultra and many 4K settings; the heaviest 4K + RT presets can still push past it later in a generation.

Best fit: Mainstream 1080p high-refresh or 1440p systems with balanced CPU and display choices.

Gaming performance

RTX 4070 Ti SUPER is a mid-range gaming GPU, at 57.81% of the dataset’s best 1440p raster performance. Match it with a 1080p high-refresh or 1440p medium-high monitor.

1440p value is 41.2 performance per dollar. Nearby GPUs in the table are often better buys.

Silicon

Full specifications

Architecture, memory subsystem, clocks, compute resources, throughput, power, and API support in one technical sheet.

Identity

GPU
RTX 4070 Ti SUPER
Chip name
AD103
Variant
AD103-275-A1
Launch price
$799
Amazon price
$1389
Availability
Unavailable
Price updated
1 week, 3 days ago
Launch
2 years, 7 months ago

Architecture

Architecture
Ada Lovelace
Foundry
TSMC
Process
5 nm
Transistors
45900 million
Die size
379 mm²
Interface
PCIe 4.0 x16

Memory

Capacity / type
16 GB GDDR6X
Bus width
256-bit
Bandwidth
672 GB/s
Memory clock
1313 MHz
L2 cache
48 MB

Clocks

Base clock
2340 MHz
Boost clock
2610 MHz

Compute

Shader cores
8448
CU / SM
56
Texture units
264
ROPs
96
RT cores
66
Tensor / matrix
264

Throughput

Pixel rate
198.0 GPixel/s
Texture rate
689.0 GTexel/s
FP32
44.10 TFLOPS
FP16
44.10 TFLOPS
FP64
689.02 TFLOPS

Power

Board power
285 W
Connectors
1x 16-pin
Reference slots
Triple-slot

APIs

DirectX
12 Ultimate (12_2)
Vulkan
1.4
OpenGL
4.6
OpenCL
3.0
Shader model
6.8

Amazon

Price history

Nearby performance

Similar GPUs

Silicon close to RTX 4070 Ti SUPER on the same 1440p scale. Open the GPU table to sort, filter, and compare side by side.

NVIDIA RTX 5070 12 GB GDDR7 · Blackwell 2.0
52.2%
45.4%
83.4
Compare these in the GPU table

Discussion

Comments on RTX 4070 Ti SUPER

No comments yet.