AMD Radeon R9 390 X2

AMD Radeon R9 390 X2

AMD Radeon R9 390 X2: A Retrospective of a Legend for Enthusiasts

April 2025

In the world of graphics cards, some models become iconic, even years after their release. The AMD Radeon R9 390 X2 is one such example. Despite being considered outdated in 2025, this card still maintains an aura of legend among hardware enthusiasts. In this article, we will explore what makes the R9 390 X2 special, how it performs today, and who might still find it useful.


1. Architecture and Key Features

Architecture: The R9 390 X2 is based on two Hawaii XT GPUs (GCN 3.0 architecture) combined using CrossFire technology. This was a solution from 2015, but in 2025 it attracts the attention of collectors and retro enthusiasts.

Manufacturing Process: The card is built on a 28nm process from GlobalFoundries. This was the standard for its time, but today such chips seem massive compared to 5nm GPUs.

Unique Features:

- FidelityFX Super Resolution (FSR): In 2021, AMD added support for FSR 1.0 for older cards, including the R9 390 X2. This allowed for an increase in FPS in games through upscaling.

- TrueAudio: Hardware audio processing — a rare feature for its time.

- Lack of Hardware Ray Tracing: Ray tracing is only implemented through software methods (e.g., in Vulkan), which severely impacts performance.


2. Memory: The Foundation of Stability

Type and Capacity: Each GPU is equipped with 8 GB of GDDR5 (a total of 16 GB, but not combined). The bus width is 512 bits for each chip.

Bandwidth: 384 GB/s for each GPU — an impressive figure even in 2025. However, GDDR5 lags behind modern HBM3 or GDDR6X in energy efficiency.

Impact on Performance: For games from 2015 to 2020, the memory capacity is sufficient even at 4K (medium settings), but in modern projects (e.g., Cyberpunk 2077: Phantom Liberty), the 8 GB per chip becomes a bottleneck.


3. Game Performance: Nostalgia in Frames

The R9 390 X2 was designed for 4K gaming, but in 2025 its niche is 1080p and 1440p.

FPS Examples (medium settings, FSR 1.0 Quality):

- The Witcher 3: Wild Hunt (1440p): 75 FPS.

- Red Dead Redemption 2 (1080p): 55 FPS.

- Hogwarts Legacy (1080p, FSR): 40 FPS (no ray tracing).

- Counter-Strike 2 (1440p): 120 FPS.

Ray Tracing: Software implementation reduces FPS by 2-3 times. For instance, in Quake II RTX, the card achieves only 25 FPS at 1080p.

Summary: For games up to 2020, it is an excellent choice. For current AAA projects, settings will need to be lowered.


4. Professional Tasks: Unexpected Longevity

Despite its gaming focus, the R9 390 X2 handles basic professional tasks:

- Video Editing: In DaVinci Resolve (OpenCL), rendering a 4K video will take twice as long as on a modern Radeon RX 7600.

- 3D Modeling: Blender (Cycles) utilizes both GPUs, but the lack of optimization for HIP (similar to CUDA) limits speed.

- Scientific Calculations: Support for OpenCL 2.0 allows the card to be used in projects like Folding@Home, but its efficiency is lower compared to NVIDIA due to the lack of CUDA.


5. Power Consumption and Heat Generation

TDP: 500 W (two GPUs at 250 W each). This requires a serious approach to cooling.

Recommendations:

- Case: Minimum 3 slots for the graphics card + good ventilation (e.g., Cooler Master HAF XB Evo).

- Cooling: Liquid cooling or top-tier air coolers (Noctua NH-D15). The card tends to overheat under load (up to 85°C).

- Power Supply: At least 750 W with an 80+ Gold rating (e.g., Corsair RM750x).


6. Comparison with Competitors

Historical Competitors (2015-2016):

- NVIDIA GTX 980 Ti: Falls short in 4K but excels in energy efficiency.

- AMD R9 Fury X: More modern HBM memory, but with less capacity (4 GB).

In 2025:

- NVIDIA RTX 4060: Consumes half the power, supports DLSS 3.5 and RTX.

- AMD RX 7600: 30% faster at 1440p, with FSR 3.0 and AV1 decoding.

Conclusion: The R9 390 X2 is a choice for those who appreciate "retro high-end" or are building PCs on older platforms.


7. Practical Tips

- Power Supply: 750 W is the minimum. It's better to have a buffer (850 W) for upgrades.

- Compatibility: Requires a motherboard with PCIe 3.0 x16 support.

- Drivers: AMD discontinued official support in 2023, but the community releases custom versions (e.g., Amernime Zone).

- Overclocking: Limited due to high TDP. It's better to focus on optimizing cooling.


8. Pros and Cons

Pros:

- Legendary status and unique design.

- High memory bandwidth.

- Support for FSR for modern games.

Cons:

- Huge power consumption.

- No hardware ray tracing or AV1 support.

- Limited driver support.


9. Final Thoughts: Who is the R9 390 X2 For?

This graphics card is an ideal option for:

- Retro gaming enthusiasts looking to build a PC from the 2010s.

- Budget builds if the card is found at a price of $150–200 (new units are rare but occasionally appear at auctions).

- Collectors who value historical models.

However, for modern tasks (games in 2025, 8K editing), the R9 390 X2 is no longer relevant. Its strength lies in nostalgia and remarkable longevity, proving that legends do not die.


Note: Prices are indicated for new devices found in limited batches or collectible editions.

Basic

Label Name
AMD
Platform
Desktop
Launch Date
September 2015
Model Name
Radeon R9 390 X2
Generation
Pirate Islands
Bus Interface
PCIe 3.0 x16
Transistors
6,200 million
Compute Units
40
TMUs
?
Texture Mapping Units (TMUs) serve as components of the GPU, which are capable of rotating, scaling, and distorting binary images, and then placing them as textures onto any plane of a given 3D model. This process is called texture mapping.
160
Foundry
TSMC
Process Size
28 nm
Architecture
GCN 2.0

Memory Specifications

Memory Size
8GB
Memory Type
GDDR5
Memory Bus
?
The memory bus width refers to the number of bits of data that the video memory can transfer within a single clock cycle. The larger the bus width, the greater the amount of data that can be transmitted instantaneously, making it one of the crucial parameters of video memory. The memory bandwidth is calculated as: Memory Bandwidth = Memory Frequency x Memory Bus Width / 8. Therefore, when the memory frequencies are similar, the memory bus width will determine the size of the memory bandwidth.
512bit
Memory Clock
1350MHz
Bandwidth
?
Memory bandwidth refers to the data transfer rate between the graphics chip and the video memory. It is measured in bytes per second, and the formula to calculate it is: memory bandwidth = working frequency × memory bus width / 8 bits.
345.6 GB/s

Theoretical Performance

Pixel Rate
?
Pixel fill rate refers to the number of pixels a graphics processing unit (GPU) can render per second, measured in MPixels/s (million pixels per second) or GPixels/s (billion pixels per second). It is the most commonly used metric to evaluate the pixel processing performance of a graphics card.
64.00 GPixel/s
Texture Rate
?
Texture fill rate refers to the number of texture map elements (texels) that a GPU can map to pixels in a single second.
160.0 GTexel/s
FP64 (double)
?
An important metric for measuring GPU performance is floating-point computing capability. Double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy, while single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable.
640.0 GFLOPS
FP32 (float)
?
An important metric for measuring GPU performance is floating-point computing capability. Single-precision floating-point numbers (32-bit) are used for common multimedia and graphics processing tasks, while double-precision floating-point numbers (64-bit) are required for scientific computing that demands a wide numeric range and high accuracy. Half-precision floating-point numbers (16-bit) are used for applications like machine learning, where lower precision is acceptable.
5.222 TFLOPS

Miscellaneous

Shading Units
?
The most fundamental processing unit is the Streaming Processor (SP), where specific instructions and tasks are executed. GPUs perform parallel computing, which means multiple SPs work simultaneously to process tasks.
2560
L1 Cache
16 KB (per CU)
L2 Cache
1024KB
TDP
580W
Vulkan Version
?
Vulkan is a cross-platform graphics and compute API by Khronos Group, offering high performance and low CPU overhead. It lets developers control the GPU directly, reduces rendering overhead, and supports multi-threading and multi-core processors.
1.2
OpenCL Version
2.0
OpenGL
4.6
DirectX
12 (12_0)
Power Connectors
4x 8-pin
Shader Model
6.3
ROPs
?
The Raster Operations Pipeline (ROPs) is primarily responsible for handling lighting and reflection calculations in games, as well as managing effects like anti-aliasing (AA), high resolution, smoke, and fire. The more demanding the anti-aliasing and lighting effects in a game, the higher the performance requirements for the ROPs; otherwise, it may result in a sharp drop in frame rate.
64
Suggested PSU
950W

Benchmarks

FP32 (float)
Score
5.222 TFLOPS

Compared to Other GPU

FP32 (float) / TFLOPS
5.59 +7%
5.432 +4%
5.147 -1.4%
5.081 -2.7%