RTX 5090 vs RTX 5080 vs RTX 5070 Ti: Which laptop GPU best deserves to top your lap?

Every mates of years, I retrieve gaming laptops exist, repressing the shame and fearfulness astatine my ain sum failures heavy capable to absorption connected the flat-folding PCs that are retired there. And, much importantly, which amongst them is worthy buying. This time, fto america see the 3 astir frame-abundant laptop GPUs successful Nvidia’s existent stable: the GeForce RTX 5090, RTX 5080, and RTX 5070 Ti, each namesakes of their variably desirable desktop counterparts, and precise apt the graphics chips du jour if you’re looking astatine the speedier extremity of the notebook market.

As ever with mobile GPUs, though, these aren’t conscionable desktop graphics cards without the blocky coolers. Much much snipping and trimming has been done to guarantee they endure the cramped innards of a portable laptop without swallowing the artillery complaint successful 1 greedy gulp or, I dunno, catching fire. The mobile RTX 5080, for instance, uses the aforesaid Blackwell architecture, offers the aforesaid afloat DLSS 4 support, and features the aforesaid 16GB of GDDR7 VRAM arsenic the desktop 5080. However! Its highest boost timepiece velocity is slowed from 2617MHz to 1500MHz, portion its CUDA halfway and RT (ray tracing) halfway counts autumn respectively from 10752 and 84 to 7680 and 60.

Image credit: Rock Paper Shotgun

The mobile RTX 5090 and RTX 5070 Ti look adjacent deeper cuts, the erstwhile lacking much than fractional the desktop version's CUDA cores (21760 versus 10496 here) and the second shrinking from 16GB of 256-bit representation to 12GB of slower 192-bit VRAM. These differences aren’t conscionable determination to shortchange laptop owners – each 3 GPUs are massively much power-efficient than the desktop engines – but they bash show, immediately, that the bigger, fatter originals aren’t reliable indicators for what you tin expect from the mobile chips.

To find retired what you can expect, I’m investigating 1 laptop apiece for each of the 3 GPUs. For the RTX 5070 Ti, it’s a Medion Erazer Major 16 X1 (£2100), and for the RTX 5080 and RTX 5090, 2 variants of the Medion Erazer Beast 16 X1 Ultimate (£2500 and £3000). Price-wise, they’re each reasonably typical, if not dead-on mean for their graphics tiers, adjacent with immoderate precise respectable physique prime and 300Hz screens; colourful and businesslike mini-LED panels, wherever the Beast 16 X1 Ultimates are concerned. Conveniently, they besides stock Intel Ultra Core 9 275HX CPUs, 2560x1600 resolutions, and 32GB RAM setups, arsenic good arsenic having been built by – presumably – the aforesaid assembly enactment robots. Any benchmarking should frankincense beryllium fairer than a crisp May morning.

Since I don’t person immoderate different GPUs tested astatine 1600p, here’s a consecutive examination of the 3 to commencement with:

Image credit: Rock Paper Shotgun

Unsurprisingly, the RTX 50 series’ wide hierarchy remains intact: the RTX 5090 is ever faster than the RTX 5080, which is successful crook ever faster than the RTX 5070 Ti. But though the 5090 hoards loads much VRAM and hundreds much of those CUDA shading cores... it’s not really that overmuch faster than the RTX 5080, is it? Just 4fps successful Total War: Warhammer III, 6fps successful some Horizon Forbidden West and F1 24, and 10fps successful Alan Wake 2 (though since each 3 mean beneath 60fps successful AW2, wherever framerate differences are much noticeable, said 10fps does astatine slightest marque for visibly smoother movements).

The summation of other tech, similar upscaling oregon ray tracing, doesn’t avail the RTX 5090 of a bully counterargument either. Horizon Forbidden West really averaged 102fps connected some the 5090 and 5080 with DLSS connected Quality mode, and though the further enabling of 2x framework procreation gave the 5090 a humble pb – 151fps to the 5080’s 143fps – erstwhile you’re moving this fast, single-digit differences are astir intolerable to spot. Cyberpunk 2077 was a akin story: with some Quality DLSS and Psycho-quality ray tracing enabled, the 5090-equipped Beast averaged 62fps, practically indistinguishable from the 5080 laptop’s 59fps astatine the aforesaid settings.

Image credit: Rock Paper Shotgun

The RTX 5070 Ti, lest we hide it, maintained a definite smoothness nether the glare of Psycho RT, producing 47fps. That’s playable, though specified a people does suggest that the RTX 5080 is inactive a wide measurement up from its next-one-down GeForce peer, successful a mode that the RTX 5090 often is not.

All these results whitethorn look adjacent much smushed-together if – yes, similar me, I cognize – you’ve antecedently lone been keeping tabs connected desktop graphics cards. In which case, you’ll callback that for arsenic overmuch of an overpriced turboindulgence the 'main' RTX 5090 is, it does astatine slightest carve retired a much commanding pb implicit different premium cards. Testing these laptops astatine 1440p and 1080p, wherever we tin commencement throwing successful comparisons betwixt desktop and mobile GPUs, illustrates conscionable however overmuch faster it is – and however large the gulf is betwixt signifier factors.

Image credit: Rock Paper Shotgun

As always, click to embiggen. | Image credit: Rock Paper Shotgun

In fairness, the laptop GPUs tin assertion immoderate extenuating circumstances, peculiarly that they’re babelike connected that (heavily efficiency-focused) Intel Ultra CPU. All the desktop paper results were taken connected the latest RPS Test Rig, with an AMD Ryzen 7 9800X3D processor that’s repeatedly proven itself the fastest gaming CPU connected the planet. That’s a large vantage adjacent earlier getting into, say, cooling differences.

Still, the extremity effect is worthy remembering: laptop GPUs aren’t conscionable a spot slower than their namesakes, they’re wholly antithetic products, and expectations should beryllium adjusted accordingly. Especially with the mobile RTX 5090, which regularly puts retired little than fractional the mean frames of the desktop model. The mobile RTX 5080 and RTX 5070 Ti besides spot overmuch little numbers, though not to anyplace adjacent the aforesaid extent. In Warhammer III, especially, the smaller, little power-hungry RTX 5080 mentation does an admirable occupation of closing the gap.

Image credit: Rock Paper Shotgun

Back connected the substance of strictly laptop vs laptop comparisons, the RTX 5080 is again lone hardly down the RTX 5090 connected average, adjacent overtaking it successful Shadow of the Tomb Raider astatine 1080p. Although it’s mean to spot narrowing framerates astatine little resolutions, wherever CPU bottlenecking becomes much of a factor, the 1440p and 1600p results besides beg the question of however it’s worthy spending hundreds much connected the 5090 for conscionable a specified sprinkling of other frames. The RTX 5070 Ti, cheaper still, holds its ain astatine these resolutions arsenic well.

The RTX 5090 doesn’t adjacent involvement a compelling assertion connected way tracing, the ray-tracing-but-more exertion that tin easy bring non-premium GPUs to their knees. PT effects should truthful beryllium the 5090’s territory, but some with and without immoderate numbers-massaging from DLSS Multi Frame Generation, the mobile RTX 5080 stays close connected its tail:

Image credit: Rock Paper Shotgun

This is besides wherever the RTX 5070 Ti makes its archetypal existent slip, being the lone 1 to autumn abbreviated of 30fps successful Alan Wake 2 without framework gen. Meanwhile, it indispensable beryllium said that the RTX 5090 does consciousness somewhat smoother conscionable from those 5 other frames implicit the RTX 5080 – an vantage that subsequently carries implicit to the 2x and 4x MFG configurations, arsenic AI frames don’t lend to the sensation of much reactive power input similar regular, rendered frames do.

Even so, of these 3 mobile GPUs, I’d spell for the RTX 5080 each time. The RTX 5070 Ti is nary slouch but successful tough-running modern games, the 5080’s show is tangibly smoother, which simultaneously makes it a amended stake for futureproofing. Conversely, the RTX 5090 doesn’t execute similar the benignant of upgrade you’d privation for an other £500. I accidental that’s little egregious than the desktop RTX 5090’s £1,000 premium implicit the desktop 5080, but past arsenic we’ve seen, the laptop 5090 lone runs fractional arsenic quickly. 50% of the show for 50% of the terms bump? It’s fitting, astatine least. Not a amended woody than the RTX 5080. But fitting.