If Nvidia’s RTX 5090 was tricky to review, its 80-class offering is even more challenging. PC users have grown used to this tier of card outperforming the outgoing flagship, but let’s be clear – RTX 5080 is a slower card than the RTX 4090 while the new RTX 5090 is much, much faster. In many ways, you can think of the new 5080 as a 4080 Super Super. There’s a small but appreciated bump to performance against its predecessor – around 13 percent in my testing – and it’s delivered for the same price. And of course, you get the latest DLSS 4 feature set, including multi frame generation, which we’ll be looking at in this review. So, this is the best GPU you can buy at its price-point – but it’s also the only new GPU available at this price-point.
Today, we’re reviewing the Founders Edition card, which is another winner from Nvidia. Up against the RTX 4080 and RTX 4080 Super, the industrial design is simply fantastic. The massive, hulking design from the last generation gives way to a new 80-class product that once again uses the 90-class cooler. That means you get a two-slot, relatively small graphics. Power is supplied by the now standard (for Nvidia) 12VHPWR socket on the top, while video outputs consist of an HDMI 2.1 port along with three DisplayPort 2.1s. I had no issues at all with noise here and similar to the RTX 5090, I loved the corrugated cardboard packaging.
Priced at $999/£979 – for the Founders model at least – I can’t help but feel that once again Nvidia has supplied the best model available at the lowest price and based on the rumours surrounding this launch (along with the extremely expensive third-party 5090s), I do have worries that the actual cost of investment in RTX 5080 may well up higher. One of the key elements that makes this product palatable is its pricing: better value than RTX 4090 and RTX 5090 in terms of ‘monetary unit per frame’ and faster/more feature-rich than RTX 4080 Super for the same price. Adding on a couple of hundred pounds or dollars brings us back into the dark days of 2022, where the RTX 4080 arrived at an unbelievable $1199/£1189.
RTX 5090 | RTX 5080 | RTX 5070 Ti | RTX 5070 | |
---|---|---|---|---|
Processor | GB202 | GB203 | GB203 | GB205 |
Cores | 21,760 | 10,752 | 8,960 | 6,144 |
Boost Clock | 2.41GHz | 2.62GHz | 2.45GHz | 2.51GHz |
Tensor Core TOPS | 3352 | 1801 | 1406 | 988 |
RT Core TFLOPS | 318 | 171 | 133 | 94 |
Memory | 32GB GDDR7 | 16GB GDDR7 | 16GB GDDR7 | 12GB GDDR7 |
Memory Bus Width | 512-bit | 256-bit | 256-bit | 192-bit |
Memory Bandwidth | 1792GB/s | 960GB/s | 896GB/s | 672GB/s |
Total Graphics Power | 575W | 360W | 300W | 250W |
PSU Recommendation | 1000W | 850W | 750W | 650W |
Power Connector | 600W PCIe 5.0 (4x 8-pin) | 450W PCIe 5.0 (3x 8-pin) | 300W PCIe 5.0 (2x 8-pin) | 300W PCIe 5.0 (2x 8-pin) |
Price | $1999/£1939 | $999/£979 | $749/£729 | $549/£539 |
Release Date | January 30th | January 30th | February | February |
Looking at specs, it’s easy to see why there is such a chasm in capabilities between RTX 5080 and RTX 5090. It’s a far cry from the RTX 30-series Ampere line, with the 3080 and 3090 were based off the same processor. In architectural terms, you can almost think of the 5080 processor as one-half of the 5090’s. The 512-bit interface becomes 256-bit (though 5080 gets slightly faster GDDR7) and there’s slightly fewer than half the CUDA cores. At 360W, the 5080 peaks at around 63 percent of the flagship’s monstrous 575W requirement. Thankfully, utilisation does not scale in a linear manner with specs, so the gaming performance gap between them closes up.
What the table doesn’t show you is how close RTX 4080 Super and RTX 5080 are – the core spec points there are very, very similar. The size of the chip is virtually identical. RTX 5080 has 10,752 CUDA cores against 4080 Super’s 10,240. Where there are variations it comes down to power and bandwidth. RTX 5080 moves up to 360W vs the 4080 Super’s 320W. The use of GDDR7 for 960GB/s of bandwidth means there’s a 30 percent uplift there compared to the outgoing card, so we should see higher than average performance uplifts on games that benefit from higher memory bandwidth. Comparisons against 4090 are also interesting – far fewer CUDA cores (10,752 vs 16,384), but slightly higher clocks and only a small reduction in bandwidth, despite the memory interface reduction. RTX 4090 will obviously be faster, but the variation in results does prove intriguing.
For all of our results, we’re using a top-end system based around the fastest gaming CPU, the AMD Ryzen 7 9800X3D, to shift the burden to the graphics card as much as possible. We also have 32GB of Corsair DDR5-6000 CL30 memory, a high-end Asus ROG Crosshair X870E Hero motherboard and a 1000W Corsair PSU.
With all that said, let’s get into the benchmarks.
Nvidia GeForce RTX 5080 Analysis
- Introduction [This Page]
- RT benchmarks: Alan Wake 2, Avatar: Frontiers of Pandora, Cyberpunk 2077
- RT benchmarks: Dying Light 2, F1 24, Hitman: World of Assassination
- RT benchmarks: Indiana Jones and the Great Circle, Metro Exodus Enhanced Edition, A Plague Tale: Requiem
- Game benchmarks: Alan Wake 2, Black Myth: Wukong, Cyberpunk 2077
- Game benchmarks: F1 24, Forza Horizon 5, Senua’s Saga: Hellblade 2
- Game benchmarks: Hitman: World of Assassination, A Plague Tale: Requiem
- DLSS 4 and Path Tracing: Cyberpunk 2077, Alan Wake 2
- Conclusions, value and recommendations
Add comment