TurboVS
AX Gaming GeForce RTX 5090 D X3W OC specifications and in-depth review

AX Gaming GeForce RTX 5090 D X3W OC

Manufacturer: AX Gaming

The AX Gaming GeForce RTX 5090 D X3W OC is a high-end graphics card occupying the top tier of the current GeForce lineup, powered by NVIDIA's Blackwell architecture. It ships with 32GB of GDDR7 video memory across a wide 512-bit bus, delivering a maximum memory bandwidth of 1790 GB/s and an effective memory speed of 28000 MHz. The card measures 358 mm in length and 148 mm in height, includes RGB lighting, and draws up to 575W under load, reflecting its position as a full-performance desktop GPU.

On the compute side, the RTX 5090 D X3W OC carries 21,760 shading units, 680 texture mapping units, and 176 render output units, producing a texture rate of 1678 GTexels/s and a pixel fill rate of 434.2 GPixel/s. Floating-point performance is rated at 107.4 TFLOPS, with a base GPU clock of 2017 MHz and a boost of 2467 MHz. The card supports DirectX 12 Ultimate, OpenGL 4.6, OpenCL 3.0, ray tracing, DLSS, and Intel Resizable BAR, while connectivity is handled through one HDMI 2.1b port and three DisplayPort outputs, supporting up to four displays simultaneously. It connects via PCIe 5.0 and is fabricated on a 5 nm process housing approximately 92.2 billion transistors.

Pros
  • The 512-bit memory bus paired with 32GB of GDDR7 VRAM and 1790 GB/s of bandwidth allows for handling large textures and complex scene data without memory bottlenecks
  • With 21,760 shading units and a boost clock of 2467 MHz, the card has substantial raw compute capacity for demanding rendering workloads
  • Support for ray tracing and DLSS together enables more realistic lighting in compatible titles while maintaining playable frame rates
  • ECC memory support makes the card viable for workloads where data accuracy matters, going beyond typical consumer graphics use
  • Four simultaneous display outputs — one HDMI 2.1b and three DisplayPort — give flexible multi-monitor setup options
  • Intel Resizable BAR support allows the CPU to access the full GPU frame buffer at once, which can reduce data transfer overhead in compatible systems
Cons
  • A TDP of 575W places significant demands on the power supply and system cooling infrastructure, requiring careful build planning
  • The card measures 358 mm in length and 148 mm in height, which may not fit in smaller or mid-tower cases
  • Air-water cooling is not included, so adequate thermal management depends entirely on the case airflow or a separately sourced cooling solution
  • There are no USB-C or DVI outputs, which limits compatibility with certain monitors or legacy display hardware without adapters
Who is this for?

This card is well-suited to users who run demanding rendering and compute workloads that benefit from 32GB of GDDR7 VRAM, high memory bandwidth, and Double Precision Floating Point support — including 3D artists, simulation engineers, and professionals working with large datasets. Enthusiast PC builders seeking maximum rasterization and ray tracing throughput will find the 21,760 shading units and 107.4 TFLOPS of compute performance appropriate for the most intensive real-time graphics tasks available today. The four-display output configuration also makes it a practical choice for users who rely on expansive multi-monitor setups for productivity or immersive simulation environments.

Who is this NOT for?

The card's 575W TDP and 358 mm length make it a poor fit for compact or small-form-factor builds where power delivery and physical clearance are constrained — users with mid-tower or smaller cases are likely to encounter compatibility issues. Those who need USB-C display output or rely on legacy DVI connections will find the port selection limiting without additional hardware. Similarly, users with basic or everyday computing needs would find the card's thermal and power demands disproportionate to their workload requirements.

Performance:

GPU clock speed 2017 MHz
GPU turbo 2467 MHz
pixel rate 434.2 GPixel/s
floating-point performance 107.4 TFLOPS
texture rate 1678 GTexels/s
GPU memory speed 1750 MHz
shading units 21760
texture mapping units (TMUs) 680
render output units (ROPs) 176
Has Double Precision Floating Point (DPFP) ✓

The Performance section of this card is defined by a base GPU clock of 2017 MHz and a boost clock of 2467 MHz, supported by 21,760 shading units, 680 texture mapping units, and 176 render output units. These figures translate into a texture rate of 1678 GTexels/s and a pixel fill rate of 434.2 GPixel/s, while floating-point performance reaches 107.4 TFLOPS. GPU memory operates at 1750 MHz, and the card includes support for Double Precision Floating Point (DPFP), making it capable of handling workloads that require higher numerical accuracy alongside its standard rendering tasks.

Memory:

effective memory speed 28000 MHz
maximum memory bandwidth 1790 GB/s
VRAM 32GB
GDDR version GDDR7
memory bus width 512-bit
Supports ECC memory ✓

This card is equipped with 32GB of GDDR7 VRAM running across a 512-bit memory bus, achieving an effective memory speed of 28,000 MHz and a maximum memory bandwidth of 1790 GB/s. ECC memory support is also included, allowing the card to detect and correct certain types of memory errors, which is particularly relevant for workloads that demand data integrity.

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

The card supports DirectX 12 Ultimate, OpenGL 4.6, and OpenCL 3, covering a broad range of modern graphics and compute APIs. Ray tracing and DLSS are both supported, along with stereoscopic 3D and multi-display technology spanning up to four screens simultaneously. Intel Resizable BAR is included to help improve data transfer between the CPU and GPU, while RGB lighting is present on the card itself. XeSS (XMX) and LHR are not featured on this model.

Ports:

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

The card's output configuration consists of one HDMI 2.1b port and three DisplayPort outputs, providing a total of four display connections. There are no USB-C, DVI, or mini DisplayPort outputs on this model.

General info:

GPU architecture Blackwell
Thermal Design Power (TDP) 575W
PCI Express (PCIe) version 5
semiconductor size 5 nm
number of transistors 92200 million
Has air-water cooling ✕
width 358 mm
height 148 mm

This card is built on the Blackwell architecture, manufactured using a 5 nm process and integrating approximately 92,200 million transistors. It connects via PCIe 5.0 and carries a Thermal Design Power of 575W, reflecting the demands of a full-performance GPU at this tier. Physically, the card measures 358 mm in width and 148 mm in height, and it does not include an air-water cooling solution.

Final Verdict

The AX Gaming GeForce RTX 5090 D X3W OC is a fully specified desktop graphics card built around NVIDIA's Blackwell architecture, combining 32GB of GDDR7 VRAM, a 512-bit memory bus, and 107.4 TFLOPS of floating-point compute performance into a single solution aimed squarely at the most demanding rendering and compute workloads available. Its support for ray tracing, DLSS, ECC memory, and four simultaneous display outputs reflects a broad feature set suited to both high-end gaming and professional application scenarios. That said, its 575W TDP and 358 mm footprint mean it demands a well-equipped, spacious system to operate properly — it is not a card for constrained builds or modest workloads. For users whose requirements genuinely justify its specification level, it represents a thoroughly capable and well-featured option within the top tier of the current graphics card landscape.