MSI Raider A18 HX review

Frame rates

1920×1080, High preset, upscaling off. Averages 182fps across the 5 core benchmark games.

Measured on this machine. Source: Notebookcheck, 2025-06-01.

In use

The short version

The MSI Raider A18 HX, in the A9WIG-080 configuration, is an 18in machine with an unusual set of priorities. It pairs an RTX 5080 held to an 85W power limit with a Ryzen 9 9955HX3D, 64GB of system memory, a 2TB solid-state drive and an 18in 3840x2400 Mini LED panel running at 120Hz. It weighs 3.64kg and carries a 99Wh battery, the largest capacity permitted in aircraft cabin baggage under standard rules.

Notebookcheck measured eleven titles on 1 June 2025, all at 1920x1080 on the High preset with upscaling switched off. The results run from 277fps in F1 24 down to 91fps in Assassin's Creed Shadows. All five games in our scoring suite are covered — Dota 2 at 271fps, GTA V at 189fps, Baldur's Gate 3 at 172fps, Cyberpunk 2077 at 168fps and X-Plane 11 at 114fps — giving a five-game mean of 183fps. That is a complete, comparable result and the machine ranks normally on our frame rate leaderboard.

There is no overall score on this page for a single reason: we have no confirmed UK price for this exact configuration. Value carries 30% of our score and cannot be calculated without one. Nothing is wrong with the machine or the data — the benchmark set is complete and stands up perfectly well. We will not guess at a price.

The central tension in this laptop is straightforward and worth stating immediately. An 85W power limit is remarkably low for an RTX 5080; it is closer to what a mainstream part is given than to what a second-tier flagship normally draws. Yet the measured frame rates are excellent — better, on this test, than a machine with an RTX 5090 at 175W. That is not a contradiction, and it is not a sign that power limits do not matter. It is a consequence of measuring at 1920x1080 with a processor that happens to be extraordinarily well suited to exactly that kind of workload. The panel, meanwhile, is 3840x2400, which is more than four times the pixel count of the resolution every one of these numbers was recorded at.

The comparison data makes the same point from the other side. Across five shared games the Raider A18 HX averages 95fps more than the MSI Cyborg 15, an RTX 5050 machine at GBP 984 — an emphatic separation. Across five shared games it averages 2fps more than the Asus ROG Strix Scar 18, which carries an RTX 5090 at 175W. Across four shared games it averages 2fps more than the Lenovo Legion Pro 7i 16 Gen 10, an RTX 5080 at 175W costing GBP 3,661. And across five shared games it averages 18fps behind the Schenker XMG Neo 16 A25 and its RTX 5090.

Buy it if you want an enormous, high-specification desktop replacement that games extremely well at 1080p and 1440p-class workloads, holds huge datasets in 64GB of memory, and has a genuinely exceptional display for content work. Buy something else if the plan is to drive that 3840x2400 panel natively in demanding games, because an 85W RTX 5080 is not the graphics implementation for that job and none of the published data suggests otherwise.

Performance

What an 85W power limit actually decides

Nvidia licenses the same graphics silicon to manufacturers across a wide band of sustained power budgets, and the chassis, cooling design and firmware decide where in that band a particular laptop settles. Two laptops can both print RTX 5080 on the lid and behave like different products under sustained load. The Raider A18 HX's figure is 85W, and that is a genuinely low number for this tier of graphics part — low enough that it should be the first thing a prospective buyer investigates rather than a footnote.

A power limit governs the clock speed a graphics card can hold once it is actually saturated. Saturation means high native resolution, heavy shader work, ray tracing, large texture working sets, and long sessions where the cooling system has stopped being able to bank heat in the metal and has to settle into a steady state. Under those conditions, the difference between 85W and 175W is not subtle. It is the difference between a graphics card running near its designed operating point and one running well below it.

A power limit decides comparatively little, however, about a graphics card that spends much of each frame idle, waiting for the processor to hand it the next batch of work. And that describes a great deal of what was measured here, because every figure in the set was recorded at 1920x1080. At that resolution, on the High preset, with no ray tracing, an RTX 5080 — even a constrained one — is ahead of the workload in most of the eleven titles listed. The bottleneck moves to single-thread processor throughput, driver overhead, memory latency and engine architecture.

This is the crux of reading this machine honestly. The 85W figure is real and it is important, and it is exactly the number that would separate this laptop from a 175W implementation at the panel's native resolution or with ray tracing enabled. It is simply not the number driving the results below. Anyone who reads a strong 1080p sweep as evidence that the power limit does not matter has drawn precisely the wrong conclusion from the right data.

The processor, and why it matters more here than usual

The Ryzen 9 9955HX3D deserves specific attention, because on this particular test it is doing an unusual amount of the work. The X3D designation denotes AMD's stacked cache implementation, which enlarges the processor's last-level cache substantially. That specific design choice benefits one class of workload above all others: games that are limited by processor throughput and memory latency rather than by graphics throughput — which is to say, games measured at 1920x1080 on a fast graphics card.

Look at the top of the measured list with that in mind. Dota 2 at 271fps is a strikingly high figure for a title whose performance profile is dominated by single-thread processor speed and cache behaviour. F1 24 at 277fps is likewise a lightly loaded, efficiently built renderer where the platform, not the graphics card, sets the ceiling. GTA V at 189fps is a decade-old engine that predates modern multi-core scheduling and is famously processor-limited at low resolutions.

These are the results that lift the five-game mean to 183fps, and they are results the processor is largely responsible for. That is not a criticism — a fast platform is a genuine benefit and produces genuinely more frames in exactly the games a lot of people play. But it does mean the headline average is describing the whole computer with a heavy weighting towards the processor, and not describing an 85W graphics card's capability at all.

Working down the eleven-game list

F1 24 at 277fps and Dota 2 at 271fps sit at the top and belong in the same category: efficiently built renderers running far ahead of anything the display can show. Racing titles designed around frame-time consistency tend to be light on the graphics card, and Dota 2 is engineered to run acceptably on modest hardware. Both figures are best read as diagnostics — the platform is working correctly, nothing is choking, drivers are behaving — rather than as performance claims. Neither is a frame rate anyone will see on a 120Hz panel.

Final Fantasy XV at 195fps and GTA V at 189fps form the next band, and both are older engines. Final Fantasy XV is a DirectX 11 title with a rendering path from an earlier era of engine design. GTA V, as noted, is heavily processor-limited under these conditions. An RTX 5080, even at 85W, is not the constraint in either.

Baldur's Gate 3 at 172fps and Cyberpunk 2077 at 168fps are the first two results where the graphics card is doing meaningful work, and they are excellent. Cyberpunk 2077 in particular remains the most demanding conventional rasterisation workload in wide benchmark circulation, and 168fps at 1080p High with upscaling off is a strong result by any standard. Baldur's Gate 3 at 172fps is comfortable enough that even the dense third act, which is known to be considerably heavier than a typical benchmark scene, should stay well clear of any threshold worth worrying about.

Indiana Jones and the Great Circle at 142fps is a modern release on a modern renderer and the first result to fall below 150fps. Still a comfortable margin over the panel's refresh rate.

Alan Wake 2 at 116fps and X-Plane 11 at 114fps land within two frames of each other for completely different reasons. Alan Wake 2 is one of the more visually ambitious modern engines in wide benchmark use and is genuinely loading the graphics card. X-Plane 11 is not loading the graphics card at all — flight simulators of that generation lean extraordinarily hard on single-thread processor performance and on the draw-call and geometry pipeline, and a laptop's X-Plane 11 result tells you a great deal about its processor and memory latency and almost nothing about its graphics card. That the two land in the same place is coincidence, not equivalence, and it is a useful reminder that position in a benchmark table does not identify the bottleneck.

Black Myth: Wukong at 96fps and Assassin's Creed Shadows at 91fps are the two lowest results and the two that matter most for judging the graphics implementation. Both are recent, graphically ambitious releases with high per-frame rendering costs, and both are firmly graphics-limited at 1080p on this machine. 91fps in Assassin's Creed Shadows at the High preset is a good result in absolute terms — smooth, comfortable, no compromise required. It is also less than a third of what the same machine produces in F1 24, and that ratio is the single most informative thing in the data set.

Where the spread comes from

The gap between top and bottom — 277fps down to 91fps — is a spread of roughly three to one, measured on one machine, on one day, at one resolution and preset. None of it comes from settings differences, thermal variation or measurement noise. All of it comes from how eleven game engines distribute their work.

The ordering is close to chronological, and that is not a coincidence. The titles at the top were built for hardware generations that no longer exist in this form, with low per-frame graphics costs; the platform runs out of processor before it runs out of graphics card. The titles at the bottom are recent releases with substantially higher per-pixel costs — denser geometry, more expensive lighting, heavier post-processing — and on those the graphics card becomes the constraint and the frame rate falls to whatever it can actually deliver at 85W.

This is where the machine's low power limit finally becomes visible in the data, and it is worth being precise about it. In the top half of the list, the 85W constraint is invisible, because the graphics card is not the limiting factor. In Black Myth: Wukong and Assassin's Creed Shadows, it is the limiting factor, and those are the two results where a 175W implementation of the same chip would be expected to pull ahead. That the machine still produces 96fps and 91fps in those titles is a testament to how capable the underlying silicon is even when constrained. It does not change the fact that it is constrained.

The 183fps average, and what it is measuring

The five-game scoring mean of 183fps is a legitimate, comparable figure, and the machine ranks normally against every other laptop measured the same way. That completeness is worth something — a fair number of machines have gaps in their suite coverage and cannot be ranked at all.

But it is worth understanding what pushes that average so high. Of the five suite games, Dota 2 at 271fps and GTA V at 189fps are both heavily processor-limited, and both are titles where the 9955HX3D's cache design is at its most effective. Those two results alone lift the mean substantially above what the graphics-limited half of the eleven-title list would suggest. Cyberpunk 2077 at 168fps, Baldur's Gate 3 at 172fps and X-Plane 11 at 114fps are the more sober members of the set.

None of that makes the average wrong. It is the correct output of a fixed, consistent test applied identically to every machine, which is precisely why the leaderboard is worth reading. But a buyer trying to predict how this laptop will handle a specific game should look at the individual figure closest to that game's engine and era, not at the mean.

What these numbers cannot tell you

There are no ray-traced figures for this machine in the data we have, no results with DLSS or frame generation enabled, and — most significantly — nothing at any resolution other than 1920x1080. On a laptop with a 3840x2400 panel, that last gap is enormous. The entire question of whether an 85W RTX 5080 can drive its own display natively is unanswered by the published measurements, and we are not going to answer it by inventing an estimate.

We also have no measured battery life, no noise figures under load, no surface temperature data, and nothing on how the machine behaves after an hour of sustained gaming. That last point is less pressing here than it would be on a 175W machine — an 85W graphics limit in an 18in chassis with a 99Wh battery and room for a substantial cooling assembly should have an easier time holding its clocks than a machine asking twice as much of the same volume — but it is an inference, not a measurement, and we will label it as such.

The screen

The panel is the most remarkable thing on the specification sheet and the source of the machine's biggest unanswered question. It is 18in, 3840x2400, 120Hz, Mini LED.

Start with the resolution, because the arithmetic is stark. 3840x2400 is 9,216,000 pixels. 1920x1080, the resolution every published frame rate on this page was recorded at, is 2,073,600. The native panel is four and a half times the pixel count of the benchmark resolution. That is not a small extrapolation gap — it is the difference between two entirely different rendering workloads, and no responsible reading of the 1080p figures can be stretched across it.

We have no measurements at 3840x2400 and will not invent any. What can be said without inventing anything is structural. The titles at the top of the list — F1 24 at 277fps, Dota 2 at 271fps — are processor-limited at 1080p, which means their frame rates would fall much less steeply with resolution than a purely graphics-limited title would. The titles at the bottom — Black Myth: Wukong at 96fps, Assassin's Creed Shadows at 91fps — are already graphics-limited at 1080p on an 85W card, and quadrupling the pixel count is exactly the change that a low power limit handles worst. The gap between those two behaviours will widen dramatically at native resolution. Where each lands is not something we can tell you from this data.

The refresh rate is 120Hz, which is a sensible pairing with a very high resolution panel — pushing much beyond that at 3840x2400 would be asking for a display specification no mobile graphics card could feed. At the measured 1080p, seven of the eleven titles exceed 120fps and would saturate the panel: F1 24, Dota 2, Final Fantasy XV, GTA V, Baldur's Gate 3, Cyberpunk 2077 and Indiana Jones and the Great Circle. Four fall short — Alan Wake 2 at 116fps, X-Plane 11 at 114fps, Black Myth: Wukong at 96fps and Assassin's Creed Shadows at 91fps — though all four remain comfortable frame rates.

The Mini LED technology is a genuine strength and a large part of what makes this machine's display worth paying for. Mini LED uses a dense array of small backlight zones with local dimming, which delivers far better contrast than a conventional edge-lit or uniformly backlit IPS panel and allows very high peak brightness, particularly in high dynamic range content. Compared with OLED it trades absolute per-pixel black levels and some blooming around bright objects on dark backgrounds for higher sustained full-screen brightness and no wear risk from static interface elements. For a machine likely to spend part of its life doing content work with fixed toolbars on screen for hours, that is arguably the more sensible technology choice.

We have no brightness measurement, no colour gamut coverage figure and no local dimming zone count for this specific panel, so we cannot quantify any of that. What is not in doubt is the shape of the proposition: this is a display specified for professional content work — high resolution, high contrast, large area — that happens to also run games at 120Hz. Reading it as a gaming panel first gets the machine wrong.

Build, size and portability

At 3.64kg the Raider A18 HX is heavy even by 18in standards, and the 18in form factor is itself the point at which a laptop stops pretending to be portable. This is a desktop replacement in the full sense: a machine that lives on a desk, that replaces a tower and a monitor, and that moves between fixed locations rather than travelling.

The practical implications are worth being blunt about. An 18in chassis does not fit an aeroplane tray table comfortably, does not fit many standard backpack sleeves, and takes up a substantial footprint on any desk it lands on. Adding 3.64kg to a bag is a decision you make deliberately, not one you make casually. And the real travelling weight is higher, because the power adapter has to come too — we do not have its weight in our data and will not put a number on the total, but nobody should budget for 3.64kg and be surprised by what actually goes into the bag.

There is a genuine benefit on the other side of that mass, and in this configuration it is an unusual one. Most 18in machines are heavy because they are cooling a high-power graphics implementation. This one is cooling an 85W graphics card in a chassis sized for considerably more. Whatever else that means, it should mean the graphics card has an easy time holding whatever clocks 85W permits, without the cooling system having to work at its limit to achieve it. We have no noise or temperature measurements to confirm that, and we are not going to present a reasonable inference as a finding, but the structural logic is sound and it is one of the more interesting things about the configuration.

The battery is 99Wh, the maximum capacity permitted in aircraft cabin baggage under standard rules, and the largest a laptop can carry without becoming a travel problem. That is the right choice for a machine of this size. We have no measured battery life figure and will not estimate one, but the structural point stands: a large cell in a machine with a relatively modest graphics power limit is a better combination for unplugged use than a large cell in a 175W machine. Gaming on battery remains something no laptop in this class does well — the graphics card cannot draw its full power from any battery and performance falls accordingly — but for non-gaming work the capacity is as good as the format allows.

The rest of the configuration is where this machine's real identity shows. 64GB of system memory is far beyond what any current game requires and is instead the specification of a workstation: large video timelines, substantial 3D scenes, many simultaneous virtual machines, big local datasets. The 2TB solid-state drive is generous enough to hold a serious game library alongside working files without constant management. And 16GB of video memory is a healthy allocation that removes texture budget worries at the resolutions this machine will realistically render at.

Put the 64GB of memory, the 2TB drive, the 3840x2400 Mini LED panel and the 85W graphics limit together and a coherent picture emerges. This is not a gaming flagship that happens to have a lot of memory. It is a large-format content machine with a fast processor and a capable-but-restrained graphics card, which also plays games extremely well. Judged as the former, the configuration makes complete sense. Judged as the latter, the 85W figure looks like an odd economy.

How it compares

Four machines from the same graphics generation, chosen to give a spread rather than just the nearest neighbours. All comparisons use only games both machines have figures for, drawn from the same 1080p High measurement conditions.

MSI Cyborg 15 — RTX 5050, GBP 984

Across five shared games the Raider A18 HX averages 95fps more than the Cyborg 15. That is the widest margin in this comparison set and by far the most important number in the section, because it establishes beyond argument that the benchmark can separate machines. A 95fps average gap across five titles is not run-to-run variance or measurement noise. It is the difference between an entry-level graphics part and a second-tier flagship, clearly and emphatically rendered by the same test that produces near-zero results elsewhere.

Keep that figure in mind for the two comparisons that follow, because the temptation when reading a 2fps gap is to conclude that the test is incapable of resolving differences. It plainly is capable. It resolved a 95fps difference here and, as we will see, an 18fps difference in the other direction. When it reports parity, it is reporting a real result about these particular games.

The Cyborg 15 costs GBP 984. We have no confirmed price for the Raider A18 HX, but an 18in machine with 64GB of memory, a 2TB drive and a 3840x2400 Mini LED panel is not competing in that bracket. The comparison is not a value judgement — it is a calibration.

Asus ROG Strix Scar 18 — RTX 5090 at 175W, no confirmed UK price

Across five shared games the Raider A18 HX averages 2fps more than the Strix Scar 18, which carries an RTX 5090 at a 175W power limit. A tier-higher graphics part, running at more than double the power budget, and the Raider is fractionally ahead.

This is the result that most needs explaining rather than repeating, because taken at face value it appears to say that graphics power limits are irrelevant. They are not. What it actually says is that in these five games, at 1920x1080, on the High preset, neither graphics card is the thing being measured for most of the run.

Three of the five suite titles — Dota 2, GTA V and X-Plane 11 — are limited by processor throughput and memory behaviour rather than by graphics throughput at this resolution. In a processor-limited title, adding graphics capability produces no additional frames whatsoever; the card finishes each frame earlier and waits longer for the next. Under those conditions an 85W RTX 5080 and a 175W RTX 5090 land in the same place, not because they are equivalent, but because neither is being asked to work. And in those specific titles the Raider has a genuine platform advantage: the 9955HX3D's cache design is aimed squarely at exactly this class of workload, which is enough to convert what should be a deficit into a two-frame lead.

Where the two machines would diverge sharply is everywhere this data does not go. At the Raider's own native 3840x2400, in ray-traced workloads, or in the heaviest current releases at raised settings, a 175W RTX 5090 has capability an 85W RTX 5080 simply does not possess. The 2fps result is real, it is correctly measured, and it applies to a narrow set of conditions that a buyer of an 18in high-resolution machine will spend relatively little of their time in.

Lenovo Legion Pro 7i 16 Gen 10 — RTX 5080 at 175W, GBP 3,661

Across four shared games the Raider A18 HX averages 2fps more than the Legion Pro 7i 16 Gen 10, which carries the same RTX 5080 at a 175W power limit and costs GBP 3,661.

This is the cleanest comparison available, because the graphics part is identical and only the power budget differs — 85W against 175W, a difference of more than two to one — and the result across four shared games is two frames in the Raider's favour. It is the most direct possible demonstration that at 1920x1080 on the High preset, a graphics power limit is very nearly irrelevant to the outcome. The graphics card is not what the test is measuring.

It is also, read the other way, the clearest possible warning about extrapolating from this test. Two identically specified graphics cards at wildly different power budgets produce the same 1080p result, which tells you the test has stopped resolving graphics capability at that point. It does not tell you the capability is absent. Move the workload to where the graphics card is genuinely the constraint — native resolution on a high-pixel-count panel, ray tracing, heavy modern engines at raised settings — and 175W against 85W becomes a substantial and measurable difference. We do not have data for that scenario on either machine, so we cannot quantify it, but the direction is not in doubt.

For a buyer choosing between them, the honest framing is this. At GBP 3,661 the Legion Pro 7i is buying the same graphics silicon with twice the power budget in a 16in package. The Raider is buying an 18in Mini LED panel at 3840x2400, 64GB of memory and a processor better suited to processor-limited games. If your gaming is at 1080p or 1440p and your other work is memory-hungry, the Raider's configuration is the more useful one. If your gaming is the heaviest current titles at native resolution, the power budget matters more than any of it.

Schenker XMG Neo 16 A25 — RTX 5090, no confirmed UK price

Across five shared games the Raider A18 HX averages 18fps behind the Schenker XMG Neo 16 A25, an RTX 5090 machine. This is the only comparison in the set where the Raider loses, and it is a useful counterweight to the two parity results above.

Eighteen frames is a modest but genuine gap — small enough that nobody would notice it in play, large enough that it is not noise. What produces it, on games that are mostly processor-limited, is almost certainly not the graphics tier alone but the whole platform: how quickly it feeds work to the graphics card, how aggressively the processor is allowed to boost, memory bandwidth and latency, and how the firmware balances power when neither component is fully loaded. The XMG Neo is a machine that consistently produces high figures on this test across a range of comparisons, which suggests a platform tuned well for exactly these conditions.

What this result does for the section as a whole is confirm that the near-zero gaps above are not the test failing. The same five games separate the Raider from the Cyborg 15 by 95fps and from the XMG Neo by 18fps. When the benchmark reports a two-frame difference against the Strix Scar 18 and the Legion Pro 7i, that is a measurement, not a shrug.

Reading the four results together

Plus 95, plus 2, plus 2, minus 18. The shape is clear: the Raider A18 HX sits comfortably in the top group of machines on this test, is not meaningfully separated from higher-power implementations within that group, and is a long way clear of the tier below.

The awkward part is that two of those three near-equals carry more graphics power than the Raider does — one with a tier-higher chip at 175W, one with the same chip at 175W. On the evidence available, the Raider's competitive position at 1080p is built substantially on its processor rather than its graphics implementation. That is a legitimate and effective way to build a fast machine for these workloads, and the numbers vindicate it. It is also a strategy whose advantage narrows as the workload becomes more graphics-bound, and the panel this machine ships with pushes the workload firmly in that direction.

Who it's for

The Raider A18 HX makes most sense for someone whose requirements are not primarily about maximum graphics throughput. If you want a large-format machine that replaces a desktop and a monitor, that holds substantial datasets in 64GB of memory, that has an 18in 3840x2400 Mini LED display for content work, and that also happens to game very well at 1080p and 1440p-class settings, this configuration is unusually well judged. The processor is excellent for the games that depend on processors, the memory and storage are generous, and the display is a serious professional tool rather than a gaming accessory.

It suits someone with a mixed workload in particular. A machine that renders video, holds large 3D scenes or runs several virtual machines during the day and plays Cyberpunk 2077 at 168fps or Baldur's Gate 3 at 172fps in the evening is doing two jobs properly rather than one job well and one badly. The 64GB of memory is not a gaming specification; it is the reason this machine can do the first job at all.

It also suits someone who has understood the comparison data correctly and finds it reassuring rather than alarming. If your gaming is at 1080p, the measured evidence is that an 85W RTX 5080 with this processor matches a 175W RTX 5090 and a 175W RTX 5080 across the shared games. For that buyer, the low power limit costs nothing they will ever notice.

Who should buy something else

If you intend to game at the panel's native 3840x2400, this is the wrong graphics implementation and the mismatch is the biggest single problem with the machine. That resolution is four and a half times the pixel count every published figure here was recorded at, and an 85W power limit is precisely the constraint that scales worst as pixel count rises. We have no native-resolution measurements and will not pretend to know where the frame rates land, but Assassin's Creed Shadows at 91fps and Black Myth: Wukong at 96fps at 1080p are already the graphics-limited end of the list. Buy a 175W machine, or plan on running games at a lower resolution than the panel's native one and using upscaling.

If ray tracing matters to you, look elsewhere or look for more data. There are no ray-traced figures for this machine at all, and ray tracing is the single workload where a constrained power budget hurts most. That is an evidence gap, not a verdict, but it is a gap in exactly the place a buyer of a high-end machine would want it filled.

If you need to carry a laptop regularly, 3.64kg plus an 18in footprint plus a power adapter is not a machine you carry — it is a machine you transport. Anyone who moves between locations more than occasionally should be looking at 16in or smaller, and accepting the smaller display and lower memory ceiling that usually comes with it.

If you want the highest measured frame rate in this comparison set, the XMG Neo 16 A25 averages 18fps ahead across five shared games. It is a smaller machine with a higher-tier graphics part, and on this test it is simply faster. We have no confirmed UK price for it, so we cannot tell you what that costs.

And if you are buying primarily on value, this page cannot help you, because there is no confirmed UK price for this configuration. What the data does give you is the reference points to judge one against when it appears: the same measured performance as a GBP 3,661 machine on four shared games, and 95fps clear of a GBP 984 one.

The bottom line

The MSI Raider A18 HX is a genuinely fast machine assembled around an unusual set of choices, and the measured data rewards it. A 183fps five-game average is a strong, complete, comparable result. Eleven titles from 277fps down to 91fps describe a laptop that handles everything currently available at 1080p without compromise. And matching a 175W RTX 5090 machine and a 175W RTX 5080 machine on the shared games is a real achievement for an 85W implementation, achieved largely through an exceptionally well-chosen processor.

The reservation is structural rather than a flaw in the measurements. This laptop ships with a 3840x2400 panel and an 85W graphics power limit, and those two specifications are pulling against each other. Every published frame rate is at a resolution using less than a quarter of the panel's pixels, so the most important performance question about the machine — can it drive its own display in demanding games — is the one the data cannot answer.

Read as a large-format content machine with strong gaming performance at sensible resolutions, it is a coherent and well-specified product. Read as an 18in gaming flagship, the power limit is the number to think hardest about, and the two parity results against 175W machines should be understood as a fact about 1080p benchmarking rather than a fact about what an 85W graphics card can do when the workload gets serious.

How it compares