Mixed- 4.5 / 5PCMag
- 70 / 100TechSpot
Efficient 4K gaming and next-gen upscaling packed into a sleek dual-slot design, held back by modest raster gains
The GeForce RTX 5080 is best suited for PC builders jumping from RTX 30-series or older cards who want smooth 4K gaming, class-leading ray tracing, and access to DLSS 4 Multi Frame Generation. Its greatest strength is its power efficiency and surprisingly slim dual-slot Founders Edition cooler, which stays cool and quiet under heavy loads. However, the generational raster uplift over the RTX 4080 Super is modest at just 8% to 11% at 4K, and retaining a 16GB memory buffer at a $999 price point limits longevity for heavier texture packs and local AI workloads. Against the Radeon RX 7900 XTX, the RTX 5080 delivers far superior ray tracing and upscaling features, but AMD offers 24GB of memory at a lower cost.
Buy it if
- Gamers upgrading from GeForce RTX 3080 or older graphics cards
- High-refresh-rate 1440p and 4K gaming with ray tracing enabled
- Compact PC builds needing a true 2-slot high-end graphics card
Skip it if
- You already own a GeForce RTX 4080 or RTX 4080 Super
- You run large local AI models or 3D rendering scenes requiring more than 16GB VRAM
- You want a massive generational leap in native rasterization performance
Key specs
- Launch price
- $999
- Reported price
- $999 to $1,150
- Architecture
- Blackwell (GB203-400)
- CUDA Cores
- 10,752 (84 SMs)
- VRAM
- 16GB GDDR7 (256-bit bus, 960 GB/s)
- Boost Clock
- 2,617 MHz (Reference)
- Rated Board Power (TGP)
- 360W
- Card Dimensions (FE)
- 304 x 137 mm, 2-slot thickness (40 mm)
- Power Connector
- 1x 16-pin 12V-2x6 (3x 8-pin adapter included)
- Display Outputs
- 3x DisplayPort 2.1b UHBR20, 1x HDMI 2.1b
In depth
Gaming performance at 1080p, 1440p and 4K
Professional testing reveals that the RTX 5080 delivers solid 4K and 1440p performance, but with unexpectedly narrow margins over the previous generation. In TechSpot's 17-game benchmark suite, the RTX 5080 averaged 91 fps at 4K, placing it 11% ahead of the RTX 4080 Super and 42% ahead of the RTX 3090. At 1440p, that advantage shrank to just 4%. Tom's Hardware reported similar results across its rasterization suite, measuring an 8.5% lead over the RTX 4080 Super at 4K and 5.2% at 1440p. At 1080p, performance essentially tied the previous generation due to platform and driver bottlenecks. In native rasterization against previous-generation flagships, TechPowerUp found the RTX 5080 lands at roughly 89% of the RTX 4090's 4K pace. GamersNexus observed individual titles such as Dying Light 2 hitting 149 fps and Final Fantasy XIV reaching 217 fps at 1440p. However, in 4K native rendering in Resident Evil 4, GamersNexus noted the 5080 fell 4 fps behind AMD's Radeon RX 7900 XTX, while trailing the RTX 5090 by nearly 69%.
Ray tracing and upscaling (DLSS/FSR)
Ray tracing performance shows a clearer architectural step forward. With 84 fourth-generation RT cores, Tom's Hardware recorded a 10% to 16% uplift over the RTX 4080 Super in heavy ray-traced workloads (including a 16% lead in Control at 4K), while outperforming AMD's RX 7900 XTX by 25% to 45%. Path tracing in intensive titles like Cyberpunk 2077 and Alan Wake 2 remains too heavy for native 4K 60 fps, requiring DLSS Super Resolution to maintain fluid frame rates. The headline addition is DLSS 4 Multi Frame Generation (MFG). In Tom's Hardware's Cyberpunk 2077 path tracing test at 4K with DLSS Performance, a 60.0 fps baseline jumped to 103.1 fps with 2X frame generation, 178.8 fps at 4X, and 247.7 fps at 6X. However, both Tom's Hardware and KitGuru pointed out that while motion appears smooth on 240Hz monitors, input polling remains anchored to the underlying rendered frame rate, resulting in around 52 ms of input latency.
Power draw, heat and noise
Despite an official 360W Total Graphics Power rating, the RTX 5080 proved remarkably frugal in real-world testing. In Tom's Hardware's gaming suite using PCAT measurement hardware, the Founders Edition averaged 297W at 4K ultra, slightly less than the RTX 4080 Super's 299W average. TechPowerUp recorded a 325W gaming average on the reference card, while GamersNexus measured desktop idle draw at 12.75W. Thermal and acoustic management on the Founders Edition earned broad praise. TechPowerUp logged GPU core temperatures of 67°C and memory temperatures of 74°C during sustained gaming, accompanied by a modest 36.8 dBA noise profile. The dual 115mm flow-through fan configuration keeps exhaust moving efficiently without requiring the massive triple-slot coolers common on third-party alternatives.
Memory (VRAM) and future-proofing
The memory configuration was the primary point of criticism across outlets. The RTX 5080 features 16GB of fast GDDR7 memory across a 256-bit bus, delivering 960 GB/s of bandwidth. While bandwidth is high, reviewers at TechSpot, KitGuru, and GamersNexus highlighted that 16GB provides no capacity increase over the RTX 4080 or RTX 4080 Super. Demanding 4K titles with full path tracing, ultra-high-resolution textures, and multi-frame generation buffers already push close to this ceiling. Furthermore, for local AI model inference, large context sizes, and complex 3D rendering scenes, 16GB falls far short of the 24GB on the older RTX 4090 or the 32GB on the RTX 5090.
Size, power connector and PSU needs
Nvidia engineered the RTX 5080 Founders Edition into a compact dual-slot form factor measuring 304 mm in length and weighing 1,643 grams. Reviewers appreciated that unlike previous 40-series behemoths and bulky partner boards, the reference card easily fits into standard mid-tower and smaller cases. Display connectivity includes three DisplayPort 2.1b ports and one HDMI 2.1b output, supporting high-bandwidth displays up to 4K at 480Hz with DSC. Power delivery relies on a single 16-pin 12V-2x6 connector. Nvidia bundles a triple 8-pin to 16-pin adapter in the box, and reviewers recommend an 850W power supply (or a quality 750W unit) to support transient spikes comfortably.
Price versus competing cards
The RTX 5080 launched at $999 MSRP, matching the starting price of the RTX 4080 Super and coming in $200 below the original launch price of the RTX 4080. In 2026, baseline models and reference editions sell near that $999 launch level, though premium partner editions range from $1,200 to over $1,600. Reviewers observed that the RTX 5080 struggles to make a compelling financial argument for owners of existing 40-series cards, given its single-digit raster gains. For buyers who want maximum performance, the RTX 5090 costs twice as much ($1,999) but offers 30% to 70% higher 4K frame rates and double the VRAM. Alternatively, AMD's Radeon RX 7900 XTX offers 24GB of VRAM and competitive 4K rasterization for $100 to $150 less, though it cannot match Nvidia's ray tracing or DLSS feature set.
Is it worth the money?
The RTX 5080 matches the $999 launch price of the RTX 4080 Super, but its modest 8% to 11% raster performance gain means it delivers incremental value rather than a generational breakthrough.