TurboVS
Inno3D GeForce RTX 4070 Ti Super X3 OC specifications and in-depth review

Inno3D GeForce RTX 4070 Ti Super X3 OC

Manufacturer: Inno3D

The Inno3D GeForce RTX 4070 Ti Super X3 OC is a graphics card built on NVIDIA's Ada Lovelace architecture, produced using a 5 nm process with 45,900 million transistors. It runs at a base clock of 2340 MHz with a boost up to 2640 MHz, and includes support for ray tracing, DLSS, stereoscopic 3D, and RGB lighting, rounding out a feature set suited to a range of rendering and compute tasks.

The card is equipped with 16GB of GDDR6X memory on a 256-bit bus, reaching an effective speed of 21,000 MHz and a maximum bandwidth of 672 GB/s. Its 8,448 shading units work alongside 264 TMUs and 96 ROPs to deliver a texture rate of 697 GTexels/s and a pixel rate of 253.4 GPixel/s. Double Precision Floating Point is supported, ECC memory is available, and the card connects via PCIe 4 with a TDP of 285W. Measuring 297 mm in width and 118 mm in height, it offers one HDMI 2.1a port and three DisplayPort outputs for up to four simultaneous displays.

Pros
  • At 297 mm in width and 118 mm in height, the card's relatively compact dimensions make it easier to fit into mid-tower and smaller cases compared to larger cards in its class
  • The 16GB of GDDR6X memory across a 256-bit bus, delivering 672 GB/s of bandwidth, supports memory-intensive rendering and compute workloads without running into VRAM constraints
  • Ray tracing and DLSS support are both present, enabling hardware-accelerated lighting and AI-driven upscaling in compatible titles and applications
  • ECC memory support adds data integrity protection useful for users running compute or professional workloads where memory accuracy is critical
  • Intel Resizable BAR is supported, allowing the CPU to access the full GPU frame buffer simultaneously on compatible platforms
  • Up to four displays can be driven at once through the combination of three DisplayPort outputs and one HDMI 2.1a port
Cons
  • Air-water cooling is not supported, which limits integration options for users who have built custom liquid cooling setups
  • There are no USB-C outputs on this card, which may be a drawback for users who rely on that connection type for displays or other devices
  • XeSS (XMX) upscaling technology is not supported, meaning users dependent on that feature will not have access to it
  • A 285W TDP still demands a well-specced power supply and adequate airflow, which may require additional planning in tightly built systems
Who is this for?

This card is well suited to users who want to run ray tracing and DLSS-enabled workloads without the physical constraints of a larger card, given its relatively compact 297 mm length and 118 mm height. The 16GB of GDDR6X VRAM with 672 GB/s of bandwidth makes it a practical choice for memory-demanding rendering tasks and compute applications, particularly those that benefit from ECC memory support. Users building systems around multi-monitor setups of up to four displays will also find the port configuration — one HDMI 2.1a and three DisplayPort outputs — a natural fit, especially on PCIe 4 platforms that support Intel Resizable BAR.

Who is this NOT for?

Users who have invested in custom liquid cooling loops will find this card incompatible, as air-water cooling is not supported, limiting thermal management options for those who prefer it. Anyone whose workflow or display configuration depends on USB-C output will need to look elsewhere, as this port type is entirely absent. Additionally, users who specifically rely on XeSS (XMX) upscaling in their applications will not find support for it here, which could be a meaningful gap depending on the software environment.

Performance:

GPU clock speed 2340 MHz
GPU turbo 2640 MHz
pixel rate 253.4 GPixel/s
floating-point performance 44.61 TFLOPS
texture rate 697 GTexels/s
GPU memory speed 1313 MHz
shading units 8448
texture mapping units (TMUs) 264
render output units (ROPs) 96
Has Double Precision Floating Point (DPFP) ✓

The GPU operates at a base clock of 2340 MHz, boosting up to 2640 MHz under load, with memory running at 1313 MHz. The 8,448 shading units are paired with 264 texture mapping units and 96 render output units, producing a texture rate of 697 GTexels/s and a pixel rate of 253.4 GPixel/s. Floating-point performance sits at 44.61 TFLOPS, and Double Precision Floating Point support is included, extending the card's utility to compute workloads alongside standard graphics rendering.

Memory:

effective memory speed 21000 MHz
maximum memory bandwidth 672 GB/s
VRAM 16GB
GDDR version GDDR6X
memory bus width 256-bit
Supports ECC memory ✓

The card features 16GB of GDDR6X VRAM running across a 256-bit memory bus at an effective speed of 21,000 MHz, translating to a maximum memory bandwidth of 672 GB/s. ECC memory support is included, providing error correction that benefits users running workloads where memory data integrity is a concern.

Features:

DirectX version DirectX 12 Ultimate
OpenGL version 4.6
OpenCL version 3
Supports multi-display technology ✓
supports ray tracing ✓
Supports 3D ✓
supports DLSS ✓
has XeSS (XMX) ✕
AMD SAM / Intel Resizable BAR Intel Resizable BAR
has LHR ✕
has RGB lighting ✓
supported displays 4

On the software and feature side, the card supports DirectX 12 Ultimate, OpenGL 4.6, and OpenCL 3, alongside hardware-accelerated ray tracing and DLSS for AI-driven upscaling. Stereoscopic 3D, multi-display technology for up to four screens, and RGB lighting are all included, while Intel Resizable BAR support is present to aid CPU-to-GPU data throughput on compatible systems. XeSS (XMX) and LHR are not supported on this model.

Ports:

has an HDMI output ✓
HDMI ports 1
HDMI version HDMI 2.1a
DisplayPort outputs 3
USB-C ports 0
DVI outputs 0
mini DisplayPort outputs 0

Display connectivity is handled through three DisplayPort outputs and one HDMI 2.1a port, totaling four available connections. USB-C, DVI, and mini DisplayPort outputs are not present on this card.

General info:

GPU architecture Ada Lovelace
Thermal Design Power (TDP) 285W
PCI Express (PCIe) version 4
semiconductor size 5 nm
number of transistors 45900 million
Has air-water cooling ✕
width 297 mm
height 118 mm

The card is based on the Ada Lovelace architecture, fabricated on a 5 nm process with 45,900 million transistors, and connects to the host system via PCIe 4. It carries a Thermal Design Power of 285W and does not support air-water cooling. Physically, the card measures 297 mm in width and 118 mm in height, giving it a relatively manageable footprint within the RTX 4070 Ti Super class.

Final Verdict

The Inno3D GeForce RTX 4070 Ti Super X3 OC presents a well-rounded specification set within the RTX 4070 Ti Super tier, combining a 2640 MHz boost clock with 16GB of GDDR6X memory delivering 672 GB/s of bandwidth — a pairing that holds up across both graphics-intensive and compute-oriented workloads. Its relatively compact 297 mm length makes system integration more straightforward than many cards at this performance level, and the inclusion of ray tracing, DLSS, ECC memory, and Intel Resizable BAR support ensures broad feature coverage. Users should be aware of the absence of USB-C output and liquid cooling compatibility when planning their build. Overall, for those on a compatible PCIe 4 platform who need a capable, moderately sized card with strong memory capacity and a full modern feature set, the Inno3D GeForce RTX 4070 Ti Super X3 OC is a solid and coherent choice.