RTX 4080 Super vs RX 7900 XTX
Features, Software, and Technology
AMD Radeon RX 7900 XTX
- FSR assistance for increased compatibility
- OpenGL 4.6, DX12 Ultimate, HDMI, and ECC Memory
- Driver advancements throughout time, despite some initial RDNA 3 issues
- Workloads with a lot of texture benefit from more VRAM.
- Not as developed as Nvidia’s software package
Nvidia GeForce RTX 4080 Super
- Frame Generation + DLSS 3
- Better support for AI software and ray tracing
- Consists of RTX Video Super Resolution, Reflex, and Nvidia Broadcast.
- A more developed ecosystem of drivers and software
- Platform of choice for creative professionals and AI developers
Gaming Performance
AMD Radeon RX 7900 XTX
- About 3% quicker at 1080p (about equal)
- At 1440p, around 4% faster (still tied)
- Approximately 7% quicker at 4K a considerable advantage in high-resolution rasterization
- Wins in some titles, such as Call of Duty: MW2 (98 fps at 4K versus 121 fps)
- A slight edge in unray-traced games’ raw gameplay performance
Nvidia GeForce RTX 4080 Super
- Some tests have a little edge at 1080p (~+3%).
- Somewhat trails at higher resolutions.
- Usually in rasterisation on par with or just below the 7900 XTX
- Some titles have more steady frame pace.
Ray Tracing Performance
AMD Radeon RX 7900 XTX
- Clearly worse performance in ray tracing
- Average loss across resolutions of 28–36%
- Far behind in ray-traced video games such as Dying Light 2, Ratchet & Clank, and Cyberpunk 2077
- When complete route tracing is enabled, the pain is increased.
- Even with upscaling, it lags since the ray tracing hardware is inferior.
Nvidia GeForce RTX 4080 Super
- Dominates ray tracing with performance that is 30–70% better.
- Ratchet & Clank is 59–74% quicker than 7900 XTX (RT on).
- Maintains a 90% lead in complete path tracing (RTX) in Minecraft.
- Keeps the lead whether upscaling is done or not.
- Much improved RT hardware and optimisation
Upscaling and AI Technologies
AMD Radeon RX 7900 XTX
- FSR 2/3 (FidelityFX Super Resolution) is supported.
- Although FSR supports all GPUs and is more open, the image quality is typically worse.
- Shaders are used instead of hardware for frame generation.
- Slower when combining upscaling and ray tracing
- Frame generation is less developed and sponsored by fewer people.
Nvidia GeForce RTX 4080 Super
- Supports the AI-powered and hardware-accelerated DLSS 3 (with frame generation).
- DLSS offers smoother performance and higher-quality images.
- Strong advantage when using DLSS + RT in games
- Fluidity is added via frame generation in addition to raw fps improvements.
- A significant edge in gaming technology for the future
Professional and AI Workloads
AMD Radeon RX 7900 XTX
- SPECviewperf 2020 performance (CAD, engineering applications) is doubled.
- More adept at rasterisation operations in Adobe Photoshop
- Could perform well in Siemens NX applications.
- Has 24GB of VRAM, which is useful for LLM work and some huge datasets.
- Training tasks and stable diffusion are limited by weaker AI acceleration.
Nvidia GeForce RTX 4080 Super
- Much quicker in testing (89%–156% faster) in AI workloads
- Outstanding results in AI training, Blender, and stable diffusion
- Utilizing the Ada Lovelace architecture, robust FP16 and FP8 computation
- Better Tensor cores and the CUDA environment for developers and AI
- Preferred for professional rendering and AI creation
Architecture and Specifications
AMD Radeon RX 7900 XTX
- Structure: RDNA 3.0
- There are 58 billion transistors.
- 5nm procedure
- 1900 MHz for the base clock and 2500 MHz for the boost
- 6144 stream processors (capable of doubling FP32)
- A total of 192 AI accelerators and 96 Ray accelerators
- 24GB of GDDR6 VRAM
- 384-bit bus width and 960 GB/s bandwidth
- Infinity Cache, 96 MB
- Power: dual/triple 8-pin connections, 355W TDP
Nvidia GeForce RTX 4080 Super
- Ada Lovelace is the architect.
- 45.9 billion transistors
- 4nm procedure
- Boost: 2550 MHz | Base Clock: 2210 MHz
- 9728 cores of CUDA
- 320 Tensor Cores | 80 RT Cores
- 16GB of GDDR6X VRAM
- Bandwidth: 736.3 GB/s | Bus Width: 256-bit
- 64 MB L2 Cache
- Power: 16-pin power connection, 320W TGP
Power Efficiency
AMD Radeon RX 7900 XTX
- 355W TDP
- Increased power consumption, particularly during load
- 96.8W for 4K YouTube playback
- 19.1W of idle power
- Stronger cooling may be required due to increased heat output.
Nvidia GeForce RTX 4080 Super
- TGP: 320W
- Much more power-efficient, particularly while working on light or idle tasks
- 31W for 4K YouTube playback
- 16.5W idle
- Reduced power means reduced heat, noise, and PSU needs.
Price
| Graphics Card | Launch MSRP (USD) | Current Street Price (Approx.) | Typical Price Range (USD) |
|---|---|---|---|
| AMD Radeon RX 7900 XTX | $999 | ~$899 – $949 | Often found below MSRP |
| NVIDIA RTX 4080 Super | $999.99 | ~$1,029 – $1,099 | Often sold above MSRP |
Availability
AMD Radeon RX 7900 XTX
- Originally more difficult to locate, now commonly accessible
- There is now a better price-to-performance ratio available.
Nvidia GeForce RTX 4080 Super
- Widely accessible, but frequently retailed for more than $1,000 over MSRP.
- More reliable availability compared to the first supply of 7900 XTX
Specifications
| Specification | AMD Radeon RX 7900 XTX | NVIDIA GeForce RTX 4080 Super |
|---|---|---|
| Architecture | RDNA 3 | Ada Lovelace |
| GPU | Navi 31 XTX | AD103-400 |
| Manufacturing Process | 5nm + 6nm (TSMC) | 4nm (TSMC) |
| CUDA / Stream Cores | 6,144 Stream Processors | 10,240 CUDA Cores |
| Ray Tracing Cores | 96 Ray Accelerators | 80 3rd Gen RT Cores |
| Tensor / AI Cores | No dedicated AI cores | 320 4th Gen Tensor Cores |
| Base Clock | ~1,900 MHz | 2,295 MHz |
| Boost Clock | Up to 2,500 MHz | Up to 2,550 MHz |
| Memory | 24 GB GDDR6 | 16 GB GDDR6X |
| Memory Bus Width | 384-bit | 256-bit |
| Memory Bandwidth | 960 GB/s | 736 GB/s |
| Total Board Power (TBP) | 355W | 320W |
| Recommended PSU | 800W | 750W |
| Power Connectors | 2 × 8-pin | 1 × 16-pin (12VHPWR) |
| PCIe Interface | PCIe 4.0 x16 | PCIe 4.0 x16 |
| Display Outputs | 2 × DisplayPort 2.1, 1 × HDMI 2.1, 1 × USB-C | 3 × DisplayPort 1.4a, 1 × HDMI 2.1a |
| Ray Tracing Performance | Moderate (less efficient than NVIDIA) | Excellent (advanced hardware support) |
| Upscaling Tech | FSR 3.0 (Frame Gen support) | DLSS 3.5 (Frame Gen, Ray Reconstruction) |
| MSRP | $999 USD | $999 USD |
| Launch Date | December 13, 2022 | January 31, 2024 |
Conclusion RTX 4080 Super vs RX 7900 XTX
AMD Radeon RX 7900 XTX
- Best value in rasterisation for expensive 4K gaming
- Additional VRAM (24GB) for professional workloads and future-proofing
- Less expensive as the RTX 4080 Super
- Perfect for those that don’t care about AI or ray tracing
- Price and high-resolution raw rasterisation wins
Nvidia GeForce RTX 4080 Super
- The best all-around player with the benefits of AI, DLSS, and ray tracing
- More feature-rich, efficient, and prepared for the future
- Significantly outperforms the RX 7900 XTX in RT and AI.
- Better for artists, broadcasters, and gamers that use state-of-the-art effects
- Gains in software, power efficiency, AI, and ray tracing