NVIDIA’s RTX 5080 Founders Edition is slim, sleek, and powerful. At least, that’s the impression you get from its marketing, but how much has changed in this redesigned card?
The “Dual Flow Through” fans are part of its shrunken design, reducing the RTX 5080 to only two slots in a standard PC case, and the adjusted 12V2x6/12VHPWR power connector should prevent any cable-melting mishaps.
Still, most of the focus is on raw performance, especially now that NVIDIA leans so heavily into AI and its upgraded DLSS 4 technology designed to upscale visuals and implement Multi-Frame Generation for smoother gameplay.
A mix of synthetic and in-game benchmarks will help compare gains and losses against previous-generation models alongside the unfortunate matter of scalpers and price gouging. So, here’s what I found in my real-world testing of the RTX 5080 FE, serving as the new high-end card for gamers who don’t need its overblown RTX 5090 sibling.
Ben Wilson
I custom-build gaming PCs for work and myself, having tested a range of high-end graphics cards and budget-grade alternatives. Everything in this review was hand-built and benchmarked in my home with standard parts.
RTX 5080: Price, availability, and specs
This review was made possible thanks to a review unit provided by NVIDIA. The company did not see the contents of this review before publishing and otherwise had no influence on the opinion.
NVIDIA’s GeForce RTX 5080 is set to launch on January 30, 2025, alongside its RTX 5090 sibling. The sale goes live at 9 AM EST in the US and 3 PM CET in the UK.
The Founders Edition (FE) GPU I’m reviewing here is available in a limited quantity, but NVIDIA’s hardware partners are filling out the selection with custom RTX 5080 GPUs. These include models from ASUS, Colorful, Galax, Gigabyte, Inno3D, MSI, PNY, ZOTAC, and more.
The RTX 5080 Founders Edition has a $999 MSRP, half as much as the RTX 5090. Custom RTX 5080 cards from partners are expected to vary in price; many will be more expensive due to extra or different cooling solutions (water-cooled cards are always pricier). It remains to be seen if there will be any custom RTX 5080 cards available for less than the FE’s $999 price tag.
Shoppers in the US can check out NVIDIA and Best Buy for the Founders Edition cards. Retailers like Newegg and B&H already have listings for third-party custom GPUs from partners to check out.
For those in the UK, NVIDIA looks to be the best place to find a Founders Edition GPU. For third-party cards, I’ve spotted listings at Overclockers UK, Scan, and Ebuyer.
👉 How to buy an RTX 5090 and RTX 5080 on launch day
With a $999 MSRP, the RTX 5080 FE comes in at $200 less than the RTX 4080 predecessor and the same price as the RTX 4080 SUPER. Here’s a look at how the raw specs of the three cards compare.
| Specs | NVIDIA RTX 5080 | NVIDIA RTX 4080 | NVIDIA RTX 4080 SUPER |
|---|---|---|---|
| CUDA cores | 10,752 | 9,728 | 10,240 |
| Shader cores | ~56 TFLOPS | 49 TFLOPS | 42 TFLOPS |
| Ray tracing cores | 171 TFLOPS | 113 TFLOPS | 121 TFLOPS |
| Tensor cores | 1,801 TOPS | 780 TOPS | 836 AI TOPS |
| Boost clock | 2.62GHz | 2.51GHz | 2.55GHz |
| Base clock | 2.3GHz | 2.21GHz | 2.29GHz |
| VRAM | 16GB GDDR7 | 16GB GDDR6X | 16GB GDDR6X |
| Memory bus | 256-bit | 256-bit | 256-bit |
| Architecture | Blackwell | Ada Lovelace | Ada Lovelace |
| TGP | 360W | 320W | 320W |
| Recommended PSU | 850W | 750W | 750W |
| Dimensions | 304mm (L) x 137mm (W) | 304mm (L) x 137mm (W) | 304mm (L) x 137mm (W) |
| Slots | 2 | 3 | 3 |
Whether or not you’re able to buy an RTX 5080 on launch day or in the weeks that follow will likely come down to luck and timing. It’s rumored that stock is extremely thin even at major retailers, with an Overclockers UK employee claiming that there are only a few hundred RTX 5080 cards available for launch.
This could, of course, change in the next couple of days, but poor timing and a short turnaround on the part of NVIDIA could have the RTX 5080 unavailable for the foreseeable future. That is, of course, excluding the scalpers who will undoubtedly be selling the new GPU for far above its suggested pricing.
RTX 5080: Who is the target audience?
The RTX 5090’s $1,999 MSRP puts it into novelty purchase territory, and there are some good reasons to avoid the card. The RTX 5080’s $999 price tag seems much more reasonable. It costs the same as the RTX 4080 SUPER at launch, and it should be a top option for any gaming enthusiasts who want a GPU that will last for years to come, even if you have a 4K monitor.
Compared to the RTX 4080 SUPER, the RTX 5080 has added more CUDA cores (10,752 vs. 10,240) and bumped up performance from the Shader, Ray Tracing, and Tensor cores. The RTX 5080 has the same 256-bit memory bus, but its 16GB of VRAM is now faster GDDR7 (compared to GDDR6X).
Cale Hunt
I’ve been building custom gaming PCs for decades, and my position at Windows Central gives me plenty of opportunity to test the latest hardware. If you need a new GPU, I can help you find the best for your build.
Whereas the RTX 5090 targets artists, scientists, AI professionals, and engineers, the RTX 5080 should be better geared for a life of gaming. It doesn’t have quite as much raw power as the RTX 4090, but the much lower price — RTX 4090 cards, if you can find them, cost between $1,500 and $2,000 — easily makes up the difference.
Compared to the RTX 4090, the RTX 5080 will offer far more TOPS — used to roughly measure AI performance — from its Tensor cores. The RTX 4090 hits 1,321 TOPS, the RTX 5090 hits 3,352 TOPS, and the RTX 5080 hits 1,801 TOPS.
Why such an emphasis on Tensor cores in the new Blackwell GPUs? Aside from the massive AI PC initiative we saw kick off in 2024, NVIDIA is relying more than ever on AI to boost gaming performance in its new GPUs.
Part of the magic comes from DLSS 4‘s new Multi-Frame Generation feature. It’s so far exclusive to the RTX 5000 GPUs — that could change in the future — and it allows for the GPU to work less while providing more frames. The RTX 5090’s massive 3,352 TOPS from its Tensor cores keeps it well in the lead, but the RTX 5080’s 1,801 TOPS is firmly in second place.
Looking at performance metrics from the RTX 5080 and RTX 5090, paying double for the latter card if you’re strictly a gamer doesn’t make much sense. The RTX 5080 comes in at about 35% below the RTX 5090 in synthetic benchmarks or about 14% below the RTX 4090.
In gaming, the RTX 5080 scores on average about 14% lower than the RTX 4090 and about 32% lower than the RTX 5090. That’s without any of the new DLSS 4 magic in place, which pushes the RTX 5080 well beyond the RTX 4090’s framerate capabilities.
Alongside the improvements to Shader, Ray Tracing, CUDA, and Tensor cores, NVIDIA has bumped the RTX 5080’s TGP up to 360W, a 40W increase compared to the RTX 4080 and RTX 4080 SUPER. That’s still 90W less than the RTX 4090.
RTX 5080: Design
Unboxing the RTX 5090 Founders Edition for my previous review set up a sense of deja vu when the RTX 5080 FE arrived since both cards are practically identical. The metal coating has an extremely subtle difference, showing an ever-so-slightly darker shade of gunmetal gray on the RTX 5090, but the two high-end cards are otherwise twins, at least to the naked eye.
Anyone opting for the RTX 5080 FE will enjoy all the benefits of NVIDIA’s generational redesign over the RTX 40 Series, including the slimmed-down GPU width and relocated power socket. Naturally, you’ll still need a PCIe 5.0 slot on your motherboard to use this card to its potential, but your power supply can breathe a sigh of relief as the power draw tops out around 360W, compared to the RTX 5090’s 575W.
Just like the RTX 5090 Founders Edition, the RTX 5080 FE comes bundled with an official NVIDIA power adapter if you don’t have a compatible 12V2x6/12VHPWR hookup on your PSU, carried over from the high-end RTX 40 Series GPUs. This time, you “only” need three 8-pin PCIe power connectors to provide the RTX 5080 with enough wattage, which can make for some messy cable management situations, depending on your case.
Either way, you shouldn’t worry too much about the risk of melting cables we saw on the RTX 4090 because NVIDIA is adamant that the new 12V2x6 pin layout is the solution to that unfortunate and dangerous issue. I have a 450W 12VHPWR cable to hand from an RTX 4090 Founders Edition, but I opted for the three-way split adapter for the sake of testing. At the minimum, I’d recommend an 850W PSU with enough PCIe connectors to handle the RTX 5080, but a reasonably priced 1000W unit would be a solid choice, too.
Personally, I use MSI’s 1300W MEG Ai1300P, but it’s over $300 and a little overkill. Testing both high-end RTX 50 Series GPUs didn’t exhibit any issues with my Ryzen 7 9800X3D build, so it’s an easy recommendation if you want to pick a PSU with guaranteed compatibility. Power draw aside, the new “Double Flow Through” fan design sees the RTX 5080 FE exhausting hot air directly upwards towards my CPU block and RAM, though my processor is somewhat protected from extra heat by ARCTIC’s Liquid Freezer III AIO block with its built-in VRM fan.
While the new power connector port sits in a far better location, the increased flow of hot air toward the top of your case means extra consideration for CPU cooling. If you opt for traditional air coolers with heatsink towers, I’d recommend monitoring your processor temperatures when pushing the RTX 50 Series cards to their limits. If your CPU does creep up its thermal metrics, it might be time to rethink your case fan layout and cooling solution.
RTX 5080: Performance benchmarks
Image 1 of 3
First up, comparing synthetic benchmark results of the RTX 5080 Founders Edition against NVIDIA’s previous-generation flagship, the RTX 4090 FE, makes for slightly more complex insights since you won’t find the latter at MSRP anywhere in 2025. Still, for the sake of data, the RTX 5080 FE at $999 MSRP performs around -14% compared to the RTX 4090 FE at $1,599 MSRP, a price tag that bordered on myths and legends for many retailers due to rampant price scalping in 2022 and beyond.
Image 1 of 3
Keeping within NVIDIA’s new RTX 50 Series generation, the RTX 5080 Founders Edition performs around -35% in these synthetic tests compared to its flagship sibling, the $1,999 RTX 5090 FE. Again, the $999 vs. $1,599 argument can only be made if a consumer has a reasonable chance of buying either card at MSRP. It might be realistic on launch day, but even then, you’d have to be lucky. After that, the crushingly disappointing reality of price gouging will likely ruin the comparison.
For a three-way stress test, I fall back on 3DMark’s Speed Way benchmark and watch the average temperatures, frame times, and power draw. Thankfully, the RTX 5080 Founders Edition runs cooler under intense usage than its RTX 5090 sibling and even slightly cooler than the chunky RTX 4090 FE flagship from NVIDIA’s previous generation. Pulling between 335W and 343W at around 67°C, the RTX 5080 runs at almost the same 2.73 GHz frequency as the RTX 4090 but operates at around a 7 FPS deficit.
The RTX 5080 FE performs at around -14% in synthetic tests compared to the RTX 4090 and -35% compared to its flagship sibling, the RTX 5090 FE.
Above all, the RTX 5080’s lower temperatures are a welcome relief, especially when compared against the last-gen RTX 4090 flagship. The slimline RTX 50 Series redesign certainly looked the part in my RTX 5090 FE review, but the higher average temperatures drew some concern, especially for a brand-new card that has years of potentially constant usage ahead of it. Still, NVIDIA removed a hotspot sensor from these new GPUs, so it’s worth checking the external temperatures before settling for benchmark metrics.
RTX 5080 temperature and fan noise
While 3DMark’s Speed Way kept running its loops in a stress test, the hottest area of the RTX 5080 FE was, predictably, in the same spot as the RTX 5090 FE but maxes out at 55.4°C instead of 66.4°C. Again, the RTX 5080 is a graphics card rated at 360W compared to the RTX 5090’s 575W, so lower temperatures are generally expected, but an 11°C reduction is a significant relief. I still believe that traditional heatsink cooling on CPUs will need extra consideration to handle this upward flow of warm air, but it’s a welcome sight.
Similarly, fan noise sees a drop from 50.5 dBA on the RTX 5090 FE to 44.9 dBA on the RTX 5080. It’s closer to my 44.7 dBA measurements of the last-gen RTX 4090 but manages it with the slimmed-down chassis of NVIDIA’s RTX 50 Series redesign. Sure, it’s definitely audible, but this is during peak performance and honestly doesn’t sound too egregious. This “Double Flow Through” setup shines more brightly with the RTX 5080 Founders Edition, even with its potential downsides regarding CPU temperatures.
RTX 5080: Gaming benchmarks
I’m using the same settings for game testing as my RTX 5090 Founders Edition review, which includes a handful of modern titles running with DLSS upscaling and frame generation disabled (except for S.T.A.L.K.E.R. 2, which has frame gen active.) Comparing first against the last-generation flagship, the RTX 4090, sees the RTX 5080 FE performing at around -14% with the highest available visual settings and 1440p resolution. 4K testing comes later, but raw performance in Baldur’s Gate 3 delivers an impressive 185 FPS, so there shouldn’t be any issue with a smooth UHD experience there.
Against the RTX 5090 Founders Edition, these benchmarks place the RTX 5080 at around -32% for gaming performance. Whether that percentage difference justifies an extra $1,000 will be a hot topic going forward, but it comes with a 575W power draw and overall higher temperatures, so it’s not without its downsides. These are also the absolute maximum settings, with many crossing a line into a placebo territory that negatively affects frame rates for often unnoticeable visual differences.
The RTX 5080 performs at around -14% in my gaming benchmarks against the last-gen RTX 4090 and -32% against its monstrous sibling, the RTX 5090.
Still, an average of 42 FPS in Cyberpunk 2077 with its Ray Tracing: Overdrive profile and Path Tracing enabled leaves a little to be desired at 1440p when the RTX 5080 should appeal as a lavish high-end GPU. The same theory applies to some of the game’s settings, and DLSS 4 goes a heck of a long way to improve this figure, as you’ll see in my next section. Nevertheless, gamers generally care about raw frames, or “brute force rendering,” as NVIDIA calls it, so this is what you can expect with the settings cranked.
Image 1 of 3
As with the synthetic benchmark tests, average temperatures remain much lower on the RTX 5080 and stay closer to the values seen on NVIDIA’s physically thicker RTX 4090 Founders Edition. A high reading of 60°C is nothing to stress about, and as long as you can push at least a steady 60 FPS, future implementations of native DLSS Multi-Frame Gen to AAA titles should be able to boost that above 120 FPS. Dropping a few settings to ‘High’ rather than ‘Epic/Ultra’ can remedy a few sub-60 FPS titles, and DLSS upscaling can take you the rest of the way to 4K if you’re adamant about maxed-out visuals.
Cyberpunk 2077: DLSS 4 MFG on vs. off
Cyberpunk 2077 is the first game to implement DLSS 4 and, with it, Multi-Frame Gen using the “Transformer” model to keep things looking clean in motion. For the most part, the implementation is as effective as it was on the RTX 5090 Founders Edition, albeit with some slightly noticeable hitches in parts. While the experience is still enjoyable overall, it still feels like a 60 FPS baseline best serves the technology, while the 45 FPS delivered by the RTX 5080 with Ray Tracing: Overdrive and Path Tracing at 1440p still earns a passing grade. Opting for 2x ironically feels smoother if you can bear a lower framerate, and the reduction of “fake frames” seems to keep things steady.
RTX 5080: Competition
It’s a little easier to point out the RTX 5080 Founder Edition’s competitors than it was for the astronomically priced RTX 5090 FE, though it’s still a continued lead for NVIDIA, no matter which way I look at it. At $999, the RTX 5080 FE could be compared against AMD’s Radeon RX 7900 XTX with its 24GB VRAM, usually selling for its intended ~$999 MSRP at Amazon. However, DLSS 4 and Multi-Frame Gen give the RTX 50 Series such a lead that AMD’s card would likely only tempt gamers away if the RTX 5080 suffers extreme price scalping, which certainly isn’t impossible.
However, even with a $1,000 difference between the RTX 5080 FE and RTX 5090, it still feels like the latter is more of a novelty despite the +32% gains in average framerates for your money. Lower overall temperatures with a cool thousand dollars sliced off the price tag still have the RTX 5080 as my top-end pick for regular enthusiasts, and the RTX 5090 doesn’t feel like it’s in the same league to act as a competitor. It’s a similar story with the RTX 4090 FE from the last gen, removing itself from the equation because of the near-impossibility of buying it at its intended MSRP.
A tempting competitor would be a previously-owned RTX 4080 SUPER or RTX 4090 at anything below $1,000.
While AMD is “taking a little extra time to optimize the (Radeon) software stack for maximum performance,” the delay of its RDNA 4 cards pushing the launch back until March 2025 means they’re of little threat to NVIDIA’s RTX 50 Series headstart. Even if some Radeon 9000 cards leak out of retailers earlier than intended, as some GPUs often do, we still have to wait for real-world comparisons before I can suggest any kind of rivalry against the advantages that DLSS 4 brings, depending on how you feel about “fake frames.”
For now, a truly tempting competitor to the RTX 5080 would be grabbing a previously-owned, or even an extremely lucky find of a brand new RTX 4080 SUPER or RTX 4090 at anything below $1,000. Some hacky solutions can potentially deliver DLSS 4 benefits to RTX 40 Series GPUs already, and NVIDIA hints at an official offering further down the line, so don’t count RTX 4070 Ti/4080/SUPER cards out just yet. Launch day is just around the corner, and its aftermath will influence competition.
RTX 5080: Should you buy it?
You should buy this if …
✅ You prefer gaming at maximum settings
If you want to crank all of the visuals in your favorite titles to their upper limits, you’ll need a powerful GPU like the RTX 5080. While not the top-end card, this will still provide a smooth experience at 4K if you utilize DLSS 4 and MFG x4.
✅ You want a slimline GPU
While AIB cards from third-party brands will offer thicker and longer RTX 5080 variants with extra fans and even overclocked internals, NVIDIA’s Founders Edition is a gorgeous two-slot offering that stays cool under stress.
You should avoid this if …
❌ You already own an RTX 4080 SUPER or 4090
The raw performance difference between the RTX 5080 and its last-gen counterparts won’t be significant enough to make a difference, especially if you prefer “brute force rendering” and want to skip upscaling and frame generation.
The RTX 5080 Founders Edition should be 60% cheaper than a previous-gen flagship MSRP, the RTX 4090 FE. However, the reality of finding the latter at anything under $2,500 would be considered extremely lucky, so it’s more like a saving of around 150%. So, while it runs at a 14% deficit in gaming, you’re saving around $1,500 for the trade-off to access DLSS 4 upscaling and frame generation benefits that combine with NVIDIA Reflex to eliminate noticeable input latency.
Even an RTX 4080 SUPER will drain your account of around $1,000, so choosing a current-gen card makes more sense, especially one with a slim redesign and modified power connectors that should, in theory, remedy the issue of melting cables. In a pure numbers game, the RTX 5080 Founders Edition comes out as the new king of the hill for sane gamers, offering a reasonable high-end without needing to spring for the absurd $1,999 MSRP of an RTX 5090 FE.
NVIDIA GeForce RTX 5080 Founders Edition
