NVIDIA GeForce RTX 5080 Review

The second graphics card launching as part of the RTX 5000 series is the GeForce RTX 5080. Hitting the shelves a week later after the arrival of the highly-anticipated RTX 5090. NVIDIA is the first computing giant to leap to GDDR7 VRAM, utilising the PCI-E 5.0 standard that has saturated the motherboard market since Intel 12th-Gen became available in 2022.

The 80-tier class of graphics cards has always been one of the most popular options within NVIDIA’s offerings, with the 1080, 2080, 3080, and 4080 being hard to come by. There are two major reasons for this. One is that these cards offer excellent performance, with the RTX 5080 targeting 4K gaming or for those building a multi-faceted workstation PC for more intense workloads. Additionally, the 80-class cards are considerably more affordable than the 90-class, making them appealing to enthusiasts who aren’t looking to break the bank.

This puts the RTX 5080 in a somewhat unique position. The new Blackwell architecture offers some significant performance boosts compared to the previous generation, along with the addition of NVIDIA’s suite of AI technologies that further enhance performance in select supported games. This card has a much more palatable MSRP of $999.99, the same price as the RTX 4080 SUPER, which landed back in 2024 as part of a mid-life refresh.

We’ll be reviewing the RTX 5080 today. We’ll explore its specs, architecture, design, and crucial performance benchmarks, which will determine whether this GPU is a worthwhile upgrade over the previous generation.

Specification

As mentioned above, the RTX 5080 has an MSRP of $999.99, the same as the RTX 4080 SUPER. It is also cheaper than the original RTX 4080, which launched at a much higher $1199.99 and wasn’t well received during initial reviews. The $1000 mark seems to be the sweet spot for top-end graphics cards, allowing gamers and enthusiasts to build a 4K-capable gaming PC for around $2000 instead of the $3000 or so that is likely to be needed for an RTX 5090 system.

The RTX 5080 boasts 16GB of GDDR7 VRAM, the same amount as previous generation options but with increased bandwidth, thanks to the fact that this is GDDR7 and not GDDR6. However, I find the 16GB VRAM slightly disappointing, especially when AMD’s competitor options, like the RX 7900 XT and XTX, offer 20GB and 24GB of VRAM, respectively. This further cements the gap between the RTX 5080 and 5090 regarding overall performance.

NVIDIA’s RTX 5080 also has a 256-bit memory bus, 10752 CUDA cores, 336 Tensor cores, and 88 RT cores. On paper, its specs are relatively similar to those of the RTX 4080 SUPER, which will be a key factor in performance improvements.

One of the considerable changes worth noting is the AI TOPS performance. AI is currently used heavily in the technology market, with much of it being buzzword nonsense in a ploy to prick up your ears. However, we wouldn’t be so quick to discard the 1801 AI TOPS figure that NVIDIA says the RTX 5080 can reach. This is so important because this is the technology used to power DLSS, and the new iteration (DLSS 4) employs a new technology known as Multi Frame Generation, which continues to provide a performance uplift at the cost of latency.

Design

The physical design of the Founders Edition card has seen multiple revisions over the years, with the RTX 3000, 4000 and 5000 series retaining similar design qualities versus their predecessors. One of the enormous changes that we’re seeing with this generation is size. Those who are frequent readers here and watchers on the GeekaWhat YouTube channel will know that we’ve had issues for many years with the sizing of RTX 3000 and RTX 4000 graphics cards. But NVIDIA has finally ended the gargantuan GPU trend by slimming down the RTX 5000 range.

Fortunately, the RTX 5090 is a two-slot card, and the RTX 5080 follows suit. Despite the higher TDP compared to the previous generation, NVIDIA has managed to shrink the size of the PCB, allowing it to sit in the middle of a much more compact chassis. The RTX 5080 utilises the same flow-through technology as the RTX 5090, channelling air to pass directly over the PCB. Cool air from your case intake will also move through the fin stack, ensuring your RTX 5080 stays nice and cool under load.

Regarding the aesthetic, the Founders Edition RTX 5080 features the same contrasting grey and black accents that we’re all used to with FE models. The sides of the GPU are rounded and feature a much lighter silver tone, while the fans and heatsink stack are darker. This provides some nice contrast but aligns with the minimalist tone that NVIDIA has set with the FE models for the past few years. Ultimately, the style likely won’t blow you away, but if you’re a fan of minimalist builds, the simplistic nature of the Founders Edition model works nicely.

Architecture

As we’ve discussed in our review of the RTX 5090, alongside the new Blackwell architecture, NVIDIA introduces an entirely new generation of DLSS, version 4.0. Similar to the RTX 4000-series with DLSS 3.0, DLSS 4.0 is an exclusive technology for the RTX 5000 range because the AI TOPS needed for the newest iteration of this upscaling and frame-gen tech is too intensive for the older cards.

To briefly explain DLSS in Layman’s terms, this technology is an AI-upscaler that uses years and years of machine learning to track and upscale frames in a video game accurately. For example, a gamer can render a game natively at a lower resolution, like 720p, and by switching on DLSS, this can be upscaled to 1080p, allowing you to enjoy the benefits of higher resolution but with minimal drawbacks.

The advantage of this is that it doesn’t require a tremendous amount of performance overhead, allowing you to enjoy a smoother framerate in your favourite games. While DLSS has been around since RTX 2000, the arrival of RTX 4000 graphics cards threw some changes into the mix with a technology known as Frame Generation. This takes data from two concurrent frames and renders an additional one in between. This increases latency but can almost double frame rates in supported titles, making it a worthwhile setting to switch on if your graphics card can’t reasonably handle the game you’re playing natively. Here are a selection of titles that support Frame Generation:

DLSS 4 improves things further by generating extra frames in a 3:1 ratio. This improvement can render three times as many for every two frames generated with frame generation.

Across our testing, we’ve found with the RTX 5090 that enabling Frame Gen 4x (the most aggressive version of this tech) framerates were nearly double compared to rasterised gameplay. In a title like Cyberpunk 2077, where ray tracing takes a massive hit to performance, turning on Frame Gen 4x is significantly worth it, allowing you to enjoy buttery smooth gameplay.

There is talk of this technology providing fake frames, and NVIDIA took some flak during its initial press conference. However, during our benchmarking, gameplay improved noticeably with Frame Gen 4x enabled compared to without. Now, this isn’t to say the tech is perfect. Firstly, it’s only supported in a handful of games, but this will increase as new titles are released over the next few years.

Additionally, this setting comes with a latency cost. It’s a bit of a mixed bag, as there are certain titles where we’d see a pretty big increase in system latency: 16ms (Cyberpunk 2077) compared to 2ms in other games (Hogwarts Legacy). In AAA titles where performance isn’t as important, this won’t matter. However, those playing FPS titles that need responsiveness will likely want to avoid this setting.

Performance

As always, all of our benchmarking and testing is done by our in-house benchmarker, Harry Coleman, who has worked tirelessly to gather all the data needed for our RTX 5000 series content. The RTX 5080 and a range of competitor graphics cards have been tested in our primary benchmarking system paired up with the AMD Ryzen 7 9800X3D to provide the best framerates possible across all of the titles we tested. The games featured in this benchmarking section provide a mix of AAA games and modern Esports titles to see how the RTX 5080 holds up in a range of games.

Cyberpunk 2077

Settings: 4K, Shadow Quality High, Indirect Lighting High, Reflections High, Crowd Density High, Particle Quality High, Volumetric Lighting High, Motion Blur off, GTAO Quality High, Grass Quality High, Contact Shadows High VSync Off, DLSS: On/Off, Ray Tracing: On/Off, Textures: High

In Cyberpunk 2077 at 4K with no additional settings, so rasterised gameplay only, the RTX 5080 offered a strong showing of 91FPS on average, with 1% lows hovering around the 63FPS mark. This is about a 15FPS gain versus the RTX 4080 SUPER, which held firm at 75FPS. In rasterised gameplay, we’re not going to see a huge performance difference, as the hardware is, for the most part, relatively similar across these two card generations, but the gain is still welcomed.

With DLSS, ray tracing, and Frame Generation 2x switched on, the RTX 5080 sees a slight gain compared to the rasterised benchmark up to 107FPS and 99FPS on the RTX 4080 SUPER. Frame Generation and DLSS are putting in some work here, but with the addition of ray tracing thrown into the mix, the performance overhead is a slight restriction.

Switching on Frame Gen 4x, we see performance skyrocket. The RTX 5080 jumps to a whopping 190FPS, leaving the RTX 4080 SUPER in the dirt. While this point of comparison isn’t entirely fair, as the RTX 4080 SUPER can’t use 4x Frame Gen, it serves as a reminder that this technology creates a pretty huge gap compared to the previous generation, which is a strong selling point for the RTX 5000 series of cards.

Alan Wake 2

Settings: 4K, Post-Processing High, Texture Resolution High, Texture Filtering High, Volumetric Lighting High, Global Illumination Quality High, Shadow Detail High, Terrain Quality High, Ray-Tracing: Enabled, Ray Tracing Preset High, DLSS: On/Off, Frame Generation: On/Off

In Alan Wake 2 at 4K high settings, rasterised only, we see the RTX 5080 output a relatively meagre 60FPS. While this is playable, it demonstrates the intensity of Alan Wake 2 and that playing without DLSS for NVIDIA cards or FSR for AMD will be a struggle. The RTX 5080, 5090, and 4090 are the few cards that can cope with this title in pure rasterised gameplay, with the RTX 4080 SUPER offering a 53FPS average, which is teetering into the realm of choppy.

Enabling Frame Gen 2x and ray tracing boosts the performance by a fair margin in Alan Wake 2. The RTX 5080 in this benchmarking run held firm at 91FPS, compared to the RTX 4080 SUPER at 83FPS. Both gameplay runs here were reasonably smooth, even with the visual fidelity increased by introducing ray tracing. Again, this test shows the strength of Frame Generation and why it’s worth turning on where available if you’re willing to sacrifice latency to a degree.

Marvel Rivals

Settings: 4K, Graphics Quality High, Global Illumination Lumen GI- High Quality, Reflection Quality Screen Space Reflections, Model Detail High, Post-Processing High, Shadow Detail High, DLSS On/Off, Frame Gen On/Off

Unsurprisingly, in Marvel Rivals, the performance of all of the graphics cards we tested was pretty strong across the board. The RTX 5080 offered a healthy 95FPS compared to the 89FPS on the RTX 4080, SUPER showing a 6FPS gain. Again, rasterised gameplay isn’t a massive indicator of generational performance improvements, especially in this Esports title, where graphics aren’t as intense as some AAA titles in our benchmarking suite.

With Frame Generation turned on, the RTX 5080 leaps to 163FPS, with the RTX 4080 SUPER sitting reasonably close behind, outputting an average of 141FPS. The framerates of the RTX 5000 series indicate that they can implement DLSS better when utilised, as they are higher than the previous generation. It’s also impressive to see how much Frame Generation uplifts performance in titles where it’s available.

COD Black Ops 6

Settings: 4K, Graphics Preset Custom, Texture Resolution High, Depth of Field On, Detail Quality High, Particle Resolution High, Shader Quality High, Shadow Quality High, Screen Space Shadows High, DLSS On/Off, Frame Gen On/Off

In COD Black Ops 6 at 4K high settings, the RTX 5080 outputs an average of 112FPSage, which is solid considering this is a rasterised benchmarking run. Again, the RTX 4080 SUPER wasn’t outpaced massively, with an average framerate of 106FPS. We did test DLSS in this benchmark, but the difference in performance is so minuscule that it isn’t worth covering in vast amounts of detail. I think the RTX 5080 performed adequately in this game, but the RTX 4080 SUPER kept up the pace.

Apex Legends

Settings: 4K, Anti-aliasing TSAA, Texture Filtering 8X, Model Detail High, VSync Disabled, Effects High, Ambient Occlusion Quality High

In Apex Legends, at 4K high, the RTX 5080 outputs an average of 266FPS, which is incredibly strong. Apex Legends is optimised, so most GPUs we test can easily handle this resolution without any problems. The RTX 4080 SUPER in this benchmark held firmly at 239 FPS, around a 30FPS loss. At framerates this high, a 30FPS loss isn’t going to make a huge difference, but it’s great to see some gains from the RTX 5080.

Fortnite

Settings: 1080p, Competitive, View Distance Far, FPS Unlimited, VSync Off, Anti-Aliasing TAA, Shadows Off, Reflections Off, Textures Low

Our last benchmarking run was Fortnite, in which we dialled the resolution down to 1080p and turned all of the settings to low. This benchmark isn’t designed to test the mettle of the cards we have access to; it provides a visual aid of generational improvements over the years. Strangely, the RTX 4080 SUPER beats the RTX 5080 with an average of 448FPS versus 440FPS. We did some retests, and the RTX 4080 SUPER just seems to perform well in this benchmark.

Conclusion