Alienware m16 R2 review

Frame rates

1920×1080, High preset, upscaling off.

Measured on this machine. Source: Notebookcheck, 2024-03-08.

In use

The short version

The Alienware m16 R2 is an early-2024 machine built around one decision, and it is the right one. The RTX 4070 inside it runs at a 140W power limit. That is the top of the band that chip is configured in, and it is the single most consequential line on the specification sheet — more consequential than the model number above it, more consequential than the panel, more consequential than anything else here.

The reason is simple and it is worth internalising before reading another word of this page. A mobile RTX 4070 is not one product. The same silicon appears in thin machines at around a third of this power budget and in mid-sized ones at roughly three quarters of it, and the difference between the bottom and the top of that range is not a rounding error. It is the difference between two laptops that share a name on the lid and behave like different tiers of hardware. Alienware have taken the ceiling of the band, and the measured data shows it.

Notebookcheck's numbers, published in March 2024 and taken at 1920x1080 on High presets with upscaling switched off, run from 255fps in Strange Brigade down to 91fps in X-Plane 11. The four games from the scoring suite that were measured average 116fps. That is a wide spread, and the shape of it is more informative than any single figure in the list: a graphics card that is being starved compresses its results, because it cannot sustain the clock speeds that light titles need in order to fly. This one does not compress. The distance between the top of the ladder and the bottom expresses the games rather than the cooling system, which is exactly what you want to see from a 140W allocation.

The complication sits on the other side of the machine. The Core Ultra 7 155H is a thin-and-light-derived processor rather than one of the HX-class desktop-derived parts that usually accompany a fully-fed 4070 in a 16in chassis, and its fingerprints are all over the lower half of the benchmark data. The lowest figure in the entire measured set is not the heaviest graphics workload. It is a flight simulator. That ordering tells you which half of this laptop runs out of headroom first, and it is not the half with the fans pointed at it.

The panel is a 16in 2560x1600 IPS running at 240Hz, and it is an ambitious pairing that the published results only partly justify. The machine weighs 2.466kg and carries a 90Wh battery, a combination that tells you a fair amount about how it is meant to be used before anyone opens the lid.

It carries no overall score. The reasons are set out in full further down, and neither of them is a criticism of the machine.

Performance

What 140W actually means

Start here, because the power limit is the most underrated number in laptop graphics and it is the one specification on this machine that decides most of the others.

Manufacturers buy graphics silicon and then choose how much sustained power to give it. That choice is made at design time, it is rarely printed on the box, and it moves performance further than any other configuration decision a laptop maker makes. A chip given a small power budget spends its life at low clock speeds under load. The same chip given a generous budget holds high clocks for as long as the cooling system can carry the heat away. Two laptops, one GPU name, results that can differ by a margin large enough to cross an entire product tier.

140W is the top of the range this part is offered in. There is no configuration of this GPU that is given meaningfully more, which means that when the benchmark data shows this machine falling behind something, the explanation is never that the graphics card was being held back. It was allowed to run. Whatever the number is, it is the number that chip produces when nothing is standing in its way.

That has a consequence for how you should read the rest of this page. On a power-restricted laptop, the interesting question about every result is how much of it was lost to the cooling system. On this one that question does not arise, so the results become a fairly clean statement about the games themselves and about the processor sitting next to the graphics card. You are reading a picture of software, not a picture of a heatsink.

Reading the shape of the spread

The published results run from 255fps at the top to 91fps at the bottom — the highest figure is close to three times the lowest. That is a very wide distribution, and width is diagnostic.

Think about what happens when a graphics card is underfed. The heaviest games in a suite are limited by how much rendering work the card can do, and they land low. The lightest games ought to land extremely high, because they ask for almost nothing per frame — but a card starved of power cannot reach the clock speeds those titles need to be pushed to their natural limits, so the top of the ladder sags towards the middle. The result is a compressed set of numbers in which everything looks samey, and that sameness is the sound of a bottleneck.

Nothing here is compressed. Strange Brigade goes where an old, extraordinarily light title should go on hardware that is not being held back. Cyberpunk 2077 goes where a demanding modern rasterised load should go. The distance between them is large because the games are genuinely far apart in cost, and the machine faithfully reports that distance rather than flattening it. This is what a healthy result set looks like, and it is the practical evidence behind the 140W figure.

The top of the ladder: 255fps, 240fps, 217fps

Strange Brigade at 255fps is the highest measured figure and the least useful for predicting what an evening with this laptop feels like. It is an old title with a low-level graphics path that has flattered every card it has ever been pointed at, and its job in a benchmark suite is diagnostic rather than representative. What 255fps tells you is that the graphics card is receiving adequate power, that nothing in the memory subsystem is choking, and that the machine can sustain very high output when a game asks for it. It is also the only title in the entire measured set that exceeds the panel's refresh rate, a point the screen section returns to.

The Witcher 3 at 240fps produces a coincidence that is almost too neat: it lands precisely on the panel's 240Hz refresh. The engine is from an earlier console generation and it is cheap to render at 1080p by 2024 standards, so this is not a stress test either — but it is a more realistic kind of light load than Strange Brigade, an actual open-world game with actual streaming and simulation behind it, and 240fps is a genuinely impressive figure for a laptop of this class. If the older half of your library looks like this, the machine has enormous headroom.

F1 22 at 217fps is the most immediately practical number at the top of the list. Racing games convert frame rate into feel more directly than almost any other genre, because your input is continuous rather than discrete and the feedback loop between hand and screen never breaks. At 217fps this machine is running a modern racing title well past the point where additional frames improve anything a human can perceive, and comfortably past the point where the panel becomes the constraint rather than the hardware. For anyone who races, this figure plus the 240Hz screen is close to an ideal pairing.

The upper middle: 150fps, 138fps, 134fps

Dota 2 at 150fps is where the first genuinely interesting thing in this data set happens, and you have to look at its neighbours to notice it.

Dota 2 has a renderer that is close to trivial at 1080p on High. It is not asking a 140W RTX 4070 to work. On machines built around desktop-derived processors, it routinely posts figures far above anything a conventional modern game manages, because the limit is not the graphics card at all — it is draw call throughput and simulation, which is to say the processor. Here it lands at 150fps, and the two titles immediately below it are Tiny Tina's Wonderlands at 138fps and Far Cry 5 at 134fps.

That is a narrow gap, and it should not be narrow. A game that barely taxes the graphics card ought to be a long way clear of two open-world titles that genuinely do. When the almost-free title sits within a handful of frames of the ones doing real rendering work, the message is that something other than the graphics card set the ceiling on the free one. That something is the Core Ultra 7 155H, and this is the first of two places in the measured data where it becomes visible.

None of which makes 150fps a bad result in absolute terms. It is a fine frame rate, it is well clear of anything you would notice, and for a competitive player it sits in the range where a high-refresh panel earns its keep. But as a measurement of hardware rather than of experience, it is modest for this class of machine, and it sets up the number at the bottom of the list.

Tiny Tina's Wonderlands at 138fps and Far Cry 5 at 134fps are both mature engines running mixed open-world loads, and both land where you would expect a well-supplied mid-to-upper mid-range card to put them. Neither is a headline result. Together they establish the machine's comfortable operating territory: a broad category of established modern games at 1080p, running with a large margin over any refresh rate a person can actually use, and with enough room left over to raise settings or push towards the panel's native resolution.

Baldur's Gate 3 at 129fps and Final Fantasy XV at 119fps

Baldur's Gate 3 at 129fps is probably the most representative single figure in the whole set, because it is a game an enormous number of people genuinely have installed and because it exercises both halves of a laptop at once. The renderer is moderate; the simulation in its busiest urban areas is not, and that combination is what makes it a good proxy for how a machine behaves when a game gets complicated rather than just prettier. At 129fps neither half of this laptop is under strain, and there is enough margin that the frame rate dips those crowded areas produce should stay well above anything uncomfortable.

It is also one of the four scoring-suite titles measured here, and it is the second-highest of them, which pulls the four-game average of 116fps upwards in a way that flatters the machine slightly relative to its two weakest results.

Final Fantasy XV at 119fps is a heavier and less forgiving graphics load than its age suggests — dense post-processing, expensive effects work, and a rendering budget that was aggressive when it arrived and has aged into a reasonable stress test. Triple-digit performance here means the card is doing sustained real work at a comfortable pace. This is the sort of title where a 4070 held to a lower power limit would start to separate from one at 140W, and the fact that it clears a hundred and nineteen frames a second with upscaling off is a straightforward point in the machine's favour.

Cyberpunk 2077 at 96fps — the graphics floor

Cyberpunk 2077 at 96fps is the yardstick, and the most honest number in the set for anyone deciding what this laptop will do with the games they are likely to buy next.

The High preset is a rasterised load. It does not switch on the ray tracing modes that reduce mid-range mobile cards to slideshows, so this is a test of conventional rendering throughput rather than of a specialised feature. At 1920x1080 with upscaling off, 96fps is a strong result for this hardware generation, and it means something specific: there is real headroom in reserve. Headroom to raise the preset. Headroom to turn on some of the more expensive lighting. Headroom to move towards the panel's native resolution and still land somewhere sensible. A machine that produces this figure at these settings is not one that will need apologising for in a demanding modern game.

It is also the lowest genuinely graphics-limited result on the page, which makes it the practical floor for anything comparably heavy. If a game you care about renders at roughly Cyberpunk's cost, this is the neighbourhood it will land in. That is a useful thing to be able to say with a measured figure behind it.

X-Plane 11 at 91fps — a processor measurement wearing a flight simulator's clothes

X-Plane 11 at 91fps is the lowest result in the measured set, and where it sits matters more than what it is.

X-Plane is close to a pure single-thread processor benchmark that happens to have aeroplanes in it. Past a fairly modest graphics threshold the card stops being the deciding factor entirely and processor throughput takes over. It is in benchmark suites precisely because it isolates the half of a laptop that most gaming benchmarks obscure.

On a machine limited by its graphics card, the bottom of the ladder is occupied by the heaviest rendering workload in the suite. That is not what happens here. Cyberpunk 2077 — the most demanding thing measured in graphics terms, running with upscaling off — finishes above the flight simulator, and only by five frames a second. A 140W RTX 4070 has been given a demanding modern title and a game with a comparatively simple renderer, and the simple one came out lower. The graphics card is not what ran out first.

Is 91fps bad? In absolute terms, no. It is a perfectly comfortable frame rate for flight simulation, a genre in which enthusiasts routinely fly happily at well under half that figure and in which scenery density matters more to most people than speed. But it establishes a ceiling, and ceilings imposed by the processor cannot be raised by lowering graphics settings. If your library leans towards flight simulation, grand strategy, large-scale city building, heavily-modded survival games or anything else where the frame rate collapses because of what the game is thinking rather than what it is drawing, this is the number to plan around — and it will not improve when you turn the shadows down.

The Core Ultra 7 155H, and why it is the interesting compromise here

Two results point in the same direction, so it is worth being explicit about the cause rather than leaving it implied.

The Core Ultra 7 155H is a mobile processor derived from the thin-and-light side of the product stack. It is not one of the HX-class parts built down from desktop silicon that appear in most 16in gaming laptops carrying a fully-powered graphics card, and it was never designed to be. Its priorities are efficiency, a modest sustained power envelope and behaviour that suits a machine that spends part of its life away from the mains. Those are sensible priorities for a portable computer. They are not the priorities that produce the highest possible frame rate in a processor-bound game.

The consequence is a laptop with an unusual internal asymmetry: the graphics card is at the absolute top of its configuration band, and the processor is emphatically not at the top of its class. In the games where the graphics card decides — Cyberpunk, Final Fantasy XV, Tiny Tina's Wonderlands, Far Cry 5, Baldur's Gate 3 — the machine performs like the 140W part it contains. In the games where the processor decides — X-Plane 11, and to a lesser but visible extent Dota 2 — it performs like the thin-and-light chip it contains.

This is not a defect, and it should not be read as one. It is a design decision with a clear logic behind it: a lower-power processor is easier to cool alongside a graphics card drawing 140W, it is kinder to a 90Wh battery when the machine is unplugged, and for the large majority of games it costs nothing measurable. It also explains a great deal about the comparison results further down this page, because at 1920x1080 the lighter titles in any shared set will be limited by the processor rather than the graphics card, and processor limits compress genuine graphics gaps into near-invisibility. When you see two machines with very different graphics cards land close together in a light title, that is usually what happened.

Where the spread comes from

Ten measured titles sort almost perfectly by what limits each frame, and once you see the ordering you can place your own library on it without needing another benchmark.

• At the very top sit old, cheap-to-render games with nothing in the way: Strange Brigade, The Witcher 3, F1 22. These say the graphics card is well-supplied and confirm the 140W figure. They say almost nothing about how the machine handles anything released recently.
• In the upper middle sit mature open-world engines doing conventional work at a conventional cost: Tiny Tina's Wonderlands, Far Cry 5. This is the machine's comfortable home territory, and a very large slice of any real games library lands here.
• Woven through both, rather than occupying a band of its own, is the processor-limited group: Dota 2 high but not as high as it should be, X-Plane 11 at the very bottom. These are the results that reveal the 155H.
• At the bottom of the graphics-limited group sits the one genuinely demanding modern rasterised load, Cyberpunk 2077, and it is close enough to a hundred frames a second to count as a good outcome rather than a warning.

The distribution is healthy. There is no cliff, no cluster of results bunched suspiciously close together, and no sign of a machine running out of thermal room in the heavier titles while the lighter ones stay artificially low. What there is, instead, is a clean split between the games that ask the graphics card questions and the games that ask the processor questions, with this laptop answering the first group better than the second.

What the published data does not contain

Three absences are worth naming rather than writing around, because on a machine like this they are not minor.

We hold no temperature measurements for this configuration. Pushing 140W through a graphics card in a 16in chassis, alongside a processor doing its own work, means a substantial quantity of heat has to leave the machine continuously. How well it does that determines whether the measured frame rates hold over a long session or sag after the first half hour, and whether the surfaces you rest your hands on stay pleasant. That is a real question and this page cannot answer it.

We hold no noise measurements either. Cooling a fully-fed 4070 is done with air moving quickly, and air moving quickly makes a sound. Machines at this power level are routinely audible under load, and the range between models is wide enough to be a genuine buying factor for anyone who plays in a shared room or with an open microphone. Again: unknown here.

And we hold no battery-runtime measurement. The 90Wh figure is a capacity, not an outcome. What it delivers in practice depends on the display's power draw at 240Hz, on how the machine behaves when unplugged, and on how aggressively it restricts the graphics card away from the mains — and 90Wh cannot feed a 140W graphics card and a processor simultaneously for any meaningful length of time, so it certainly does restrict it. How much, and what that leaves you with, is not something the published data settles.

Those three gaps do not change the frame rate analysis above, all of which stands on measured figures. They do mean that a chunk of the ownership experience on a high-power machine is genuinely unquantified here, and it would be dishonest to fill it with confident guesses.

The screen

240Hz on a 1600p panel is an aggressive pairing

A 16in 2560x1600 IPS panel at 240Hz is a specification that reads beautifully and needs examining carefully, because the two headline numbers pull against each other. High resolution costs frames. High refresh needs frames. Putting both on the same panel means the hardware behind it has to satisfy two demands that are in direct tension.

Take the refresh rate first, measured against the published results. Every figure on this page was recorded at 1920x1080. At that resolution, exactly one measured title exceeds 240fps: Strange Brigade, at 255fps. One more lands on the number precisely: The Witcher 3, at 240fps. F1 22 at 217fps approaches it closely enough that a variable refresh implementation would keep the two in step comfortably. Everything else — every scoring-suite title, every modern game, the flight simulator, the two open-world titles — sits below the panel's ceiling, and the four-game scoring average of 116fps is less than half of it.

That is not a criticism, and it is important not to turn it into one. A 240Hz panel is not wasted at a hundred and thirty frames a second. Motion clarity improves with refresh rate independently of frame rate, input latency falls, and a panel with headroom above your typical frame rate is a panel that never becomes the constraint. The honest framing is this: the 240Hz figure is fully exploited only by old, light titles, and everything else benefits from the panel partially rather than completely. If you buy this machine specifically to run modern games at the refresh rate on the box, the measured data does not support that expectation.

One caveat that cuts both ways: we have no confirmation of adaptive refresh support for this exact configuration. On a machine whose frame rates mostly sit well below the panel's ceiling, whether the display can vary its refresh to match the output matters a great deal to how smooth those frame rates feel. It is a significant unknown rather than a footnote.

What happens at 2560x1600, and why this section contains no frame rates

Every published figure quoted on this page is at 1920x1080. There are no measurements at the panel's native resolution, so anything said about native performance is inference, and it will be labelled as such rather than dressed up as data.

The arithmetic is straightforward. 2560x1600 contains very close to twice as many pixels as 1920x1080 — just under double, and near enough to double that you should treat it as double when reasoning about it. That is a substantial increase in the work the graphics card does per frame.

What it is not is a straightforward halving of frame rate, and the reason is the split established in the performance section. In a title limited by the graphics card, doubling the pixel count moves the frame rate a long way down, though generally by less than the pixel ratio because a portion of every frame's cost does not scale with resolution. In a title limited by the processor, doubling the pixel count barely moves anything at all, because pixels were never the constraint. Dota 2 and X-Plane 11 would lose comparatively little at native resolution. Cyberpunk 2077 and Final Fantasy XV would lose a great deal.

The practical implication is that the ladder reorders itself at native resolution. The processor-limited titles climb relative to everything else; the graphics-limited ones fall; and the gap between the top of the list and the bottom narrows considerably. That reordering is a reliable inference from the way the two groups behave. The specific numbers on the other side of it are not available, and this page will not invent them.

What can be said with confidence is that nothing in the measured set would approach 240fps at 2560x1600. Strange Brigade is the only title that clears the panel's refresh at 1080p, and it clears it by a modest margin. Doubling its pixel workload would put it well underneath. At native resolution, the 240Hz panel is a specification you own rather than one you use, unless you are prepared to lower presets substantially, enable upscaling, or play games considerably lighter than anything in this benchmark set.

Upscaling deserves a fair hearing here, because it is the obvious answer and it is a legitimate one. The measured figures were deliberately recorded with upscaling off, which is correct methodology for comparing hardware and is not how most people actually play. Rendering at a lower internal resolution and reconstructing to 2560x1600 is precisely the tool this situation calls for, it is well-supported on this generation of hardware, and on a 16in panel the reconstruction artefacts are far less visible than on a large desktop display. The realistic way to use this machine is at native resolution with upscaling engaged, and there is nothing shameful about that. It is simply worth understanding in advance rather than discovering afterwards.

IPS rather than OLED

The panel is IPS, which in 2024 was a deliberate choice rather than a cost-saving default, because OLED alternatives existed in this size and price bracket — one of the rivals below is an OLED machine.

The trade is well understood. OLED gives near-perfect blacks, effectively unlimited contrast and superb colour reproduction. IPS gives consistent, predictable output with no risk of static elements gradually burning into the display, and on a machine whose owner will spend hundreds of hours looking at the same game HUD in the same position, that is a real consideration rather than a theoretical one. A gaming laptop is also a laptop, which means taskbars, browser tabs and application chrome sitting in fixed positions for years.

We have no measured brightness or colour figures for this configuration, so this section cannot tell you whether this particular IPS panel is a good one. It can only tell you what the technology choice implies. For a machine expected to be used daily for mixed work and games, IPS at this resolution and refresh is a defensible and arguably sensible selection, and it pairs logically with the 240Hz target — high-refresh IPS panels of this era were mature and widely available, whereas high-refresh OLED was newer and less consistent.

Build, size and portability

2.466kg, and what the weight is buying

The published specifications give a weight of 2.466kg and a battery of 90Wh. They do not give dimensions or thickness, so this section will not pretend to know them. There is still a great deal that can be reasoned out from the numbers that do exist, because weight on a gaming laptop is not an accident — it is the physical cost of the other decisions on the specification sheet.

Nearly two and a half kilograms in a 16in machine is substantial without being extreme for its class. It is not a laptop you forget about in a bag. Add a mains adapter capable of feeding a system with a 140W graphics card in it — and no adapter figure is published here, so its exact rating is unknown, but it will not be small — and the travelling weight climbs further. This is a machine you carry deliberately, on journeys you have planned for, not one you sling into a rucksack on the way out of the door.

What that mass is buying is precisely the thing this review has spent most of its length praising. Sustaining 140W of graphics power requires heatsink volume, heatpipe or vapour chamber area, fans of a usable diameter, and a chassis rigid enough to hold all of it without flexing. Those things are heavy. A thinner, lighter machine with the same graphics card badge on it does not have room for that hardware, which is exactly why thinner machines with the same badge are configured at lower power limits. The weight and the frame rates are two views of the same engineering decision.

This is worth stating plainly because the market presents it as a menu of independent choices, and it is not. You cannot select a 4070 at 140W and also select a light chassis. The relationship between the two is not marketing; it is thermodynamics. If the frame rates in the performance section are what you want, 2.466kg is roughly what they cost.

The 90Wh battery

90Wh is a large laptop battery, close to the upper limit of what can be carried in aircraft cabin baggage, and it is the specification that most clearly signals the machine's intended split between desk and elsewhere.

It cannot, however, do what the machine can do on mains power. A graphics card configured at 140W, running alongside a processor and a 240Hz display, draws more than a 90Wh pack can supply for any usable length of time. Every gaming laptop restricts its graphics card away from the mains as a result, and this one certainly will. Since we have no runtime measurement for this configuration, this page cannot tell you what unplugged gaming looks like here, and it will not guess.

What the battery does buy is genuine non-gaming endurance, and this is where the Core Ultra 7 155H stops being a compromise and becomes an asset. An efficiency-oriented processor with a modest sustained power envelope, paired with a large battery, is a much better foundation for a day of browser tabs and documents than a desktop-derived chip would be. The processor choice that costs frames in X-Plane 11 is the same choice that makes this machine plausible as a computer rather than purely as a games console with a keyboard attached. That is a coherent design, even if it is not the one a pure frame rate maximiser would have made.

16GB of memory

16GB was the working floor for a gaming laptop in 2024 rather than a comfortable allowance, and pairing it with a graphics card at the top of its power band is a slightly odd balance. The graphics side of this machine has been given everything available to it. The memory has been given the minimum that the class expects.

In the measured titles this will rarely be the constraint, and none of the results on this page look memory-starved. The concern is forward-looking rather than immediate: system memory requirements have risen steadily, background applications have grown, and a machine whose graphics card will still be perfectly capable in several years may find its memory the thing that dates first. Whether that can be addressed depends on whether the modules are socketed, which the published specifications here do not state, so it is a question to resolve before purchase rather than after.

How it compares

Four machines from the same graphics generation, compared only on games both laptops hold, with preference given to the scoring-suite titles. Each comparison is an average across the shared set, so the number of shared games matters as much as the size of the gap — a seven-game comparison is a far more robust statement than a four-game one.

Gigabyte Aero 14 OLED BMF — the machine it beats by most

Across the six games these two share, the m16 R2 averages 79fps ahead. That is an enormous margin, and it is the cleanest illustration on this page of what a real hardware gap looks like.

The Aero carries an RTX 4050. That is two steps down the product stack, and in a 14in chassis it will also be operating in a very different power envelope from 140W. The two effects stack: a smaller chip, given less power, in a machine with less room to remove heat. Averaging 79fps behind across six titles is not a disappointing showing for the Aero — it is the arithmetic of the category it belongs to.

The useful conclusion is not that one machine is better. It is that these two are not really competing. The Aero is a 14in OLED laptop for someone whose first requirement is that the machine travels easily and whose gaming is a secondary function. The m16 R2 is for someone whose first requirement is frame rate. If you are genuinely torn between them, the frame rate data has already told you which requirement is your real one, because a gap that size is not something a preference for a nicer screen can absorb.

Aorus 15X ASF — the confirmation

Across the seven games these two share, the average difference rounds to nothing. Not close: nothing. And this is the most quietly informative comparison on the page.

The Aorus 15X ASF carries an RTX 4070 at 140W. Same chip. Same power limit. Two different manufacturers, two different chassis, two different cooling designs, two different sizes of machine — and across seven shared titles the frame rates land on top of each other.

It would be easy to describe that as a null result and move on. It is the opposite. This is a controlled experiment appearing by accident in a comparison table, and it validates the argument this entire review is built on. If the graphics card model determined performance on its own, machines with the same card at different power limits would perform the same, and they conspicuously do not. If the chassis and cooling determined performance, two 140W machines from different manufacturers in different sizes would diverge, and they conspicuously do not either. What actually happened is that two laptops running the same silicon at the same sustained power produced the same frame rates, which is exactly what the theory predicts and exactly what buyers are rarely shown.

The practical advice that falls out of this is unusually concrete. If you are choosing between these two machines, performance is not a factor, because there isn't any difference to weigh. Decide on the screen, the size, the keyboard, the memory configuration, the warranty and the price, and treat the frame rates as a solved question. That is a genuinely useful thing to be able to tell someone.

HP Omen Transcend 16-u1095ng — close, on a thinner basis

The m16 R2 averages 1fps ahead across the four games these two share. Another RTX 4070, another result that lands within a single frame — but this one deserves a caveat that the Aorus comparison does not.

Four shared titles is a thin foundation. With seven games, a result that lands level is a strong statement, because the set is broad enough to include titles limited by the graphics card, titles limited by the processor and titles limited by neither. With four, a single unrepresentative pairing carries much more weight in the average, and if the shared four happen to skew towards lighter or more processor-bound titles, genuine graphics differences get compressed out of the result. That compression is real and it is the mechanism described in the performance section: at 1920x1080, light titles are decided by the processor, and processors hide graphics gaps.

So the honest reading is that these two machines look extremely close on the evidence available, and that the evidence available is narrower than one would like. The Transcend is a slimmer 16in design, which raises an obvious question about how its 4070 is configured; if it runs at a lower power limit than 140W, you might expect a gap to open up in heavier titles that this four-game set does not fully cover. The published data does not settle that, and this page will not pretend otherwise.

Asus ROG Strix Scar 17 — the machine that beats it by most

Across the seven games these two share, the m16 R2 averages 110fps behind the Scar 17. That is a very large margin in the other direction, and it puts the m16 R2's position in the generation into perspective.

The Scar 17 is an RTX 4090 laptop — the top of the generation's mobile stack, in a 17in chassis with the cooling capacity and the weight that implies. A gap of that size across seven titles is what the flagship tier looks like when it is given room to work. There is no configuration argument to be made here and no consolation to offer: the m16 R2 is comprehensively outperformed, and the only surprise would have been if it were not.

What matters is that these two are not alternatives to one another in any practical sense. A 17in flagship is a larger, heavier machine with a different cooling system and, inevitably, a different price. The comparison is useful for calibration — it shows you how much performance exists above this machine if you are willing to carry and pay for it — rather than as a purchasing decision. A well-fed 4070 is a strong upper-mid-range proposition for its era. It was never the top of the stack, and 110fps is roughly what the distance to the top of the stack was worth in this generation.

What the four comparisons say together

Line them up and a single principle explains all four outcomes. Against a card two tiers below in a thin chassis: an enormous win. Against the identical card at the identical power limit: no difference at all. Against another card of the same model in a slimmer machine: within a frame, on a limited sample. Against the generation's flagship: a large, unambiguous loss.

Nothing in that pattern is explained by brand, by chassis design or by cooling cleverness. All of it is explained by which chip is inside and how many watts it is allowed to draw. That is a slightly deflating conclusion for anyone who enjoys reading about heat pipe layouts, but it is what the measured data across four rivals actually shows, and it is the single most useful thing a buyer can take away from this page.

Why there is no score

Two reasons, both about the data rather than the machine.

The first is the benchmark suite. Every laptop that carries a frame rate ranking on this site is averaged over the same five scoring-suite games, so that the comparison between any two machines is genuinely like for like. For the m16 R2, four of those five were measured — Dota 2 at 150fps, Baldur's Gate 3 at 129fps, Cyberpunk 2077 at 96fps and X-Plane 11 at 91fps, averaging 116fps — and there is no figure for GTA V. A four-game average cannot be placed on a ladder built from five-game averages without quietly changing what the ladder means, particularly when the missing title is an old, processor-dependent one that would interact with this machine's specific strengths and weaknesses in a way none of the other four do. That is a gap in the data, not a judgement on the laptop, and the four figures that do exist are analysed in full above.

The second is the price. We have no confirmed UK price for this configuration, and value is a substantial component of any score worth publishing, so no value judgement is possible and none will be offered — no estimate, no range, no guess at what it ought to cost. A machine that is excellent at one price is poor at another, and speculating would give you a number that looks authoritative and means nothing.

Who it's for

Buy it if

You want an RTX 4070 that behaves like an RTX 4070. This is the central case for the machine and it is a strong one. The 140W configuration is the top of the band, the measured results confirm it across ten titles, and the comparison against an identically-configured rival confirms that the number rather than the badge is what produced them. Buying a 4070 in a thinner machine means buying the same name and a fraction of the output.

Your games are demanding to render rather than demanding to simulate. Modern rasterised titles are where this laptop is at its best: Cyberpunk 2077 at 96fps with upscaling off is a genuinely comfortable result, and the mature open-world titles sit far above anything you would notice. If your library is mostly shooters, action games, racing and modern single-player releases, the performance section describes a machine that will handle them well.

You play at 1080p, or at native resolution with upscaling engaged. Both are entirely reasonable ways to use this screen, and the second is what most owners will actually do. Approached that way, the resolution is an asset for everything that is not a game and no real obstacle to the games themselves.

The machine will mostly live on a desk and travel occasionally. At 2.466kg with a large adapter, that is the usage pattern the hardware suits, and the 90Wh battery plus efficient processor makes the occasional untethered day genuinely workable for non-gaming tasks.

Buy something else if

Your games are processor-bound. This is the clearest exclusion and the measured data states it plainly. X-Plane 11 at 91fps is the lowest figure in the set, below a demanding modern shooter running with upscaling off, and Dota 2 at 150fps sits closer to the graphics-limited titles than it should. Flight simulation, grand strategy, large city builders, heavily-modded sandboxes and simulation-heavy games in general will find the Core Ultra 7 155H before they find the graphics card, and no settings change will recover that. A machine with an HX-class processor is the right answer for that library, even if it carries the same graphics card.

You need to carry it every day. 2.466kg plus an adapter for a system with a 140W graphics card is a considered journey, not a casual one. If portability is a genuine daily requirement rather than an occasional one, a lighter machine with a lower-powered graphics card — the Aero 14 OLED being exactly that proposition — will serve you better, and the 79fps you give up is the honest price of that decision.

You want to run modern games at 2560x1600 without upscaling. The published results are all at 1920x1080 and there is no native-resolution measurement here, but roughly doubling the pixel count of a title already at 96fps does not leave a great deal of margin. Upscaling is the answer and it is a good one; if you object to it on principle, this is not the machine for you.

You are choosing between this and the Aorus 15X ASF on performance grounds. There are no performance grounds. Across seven shared titles the difference rounds to nothing, so whichever of the two is better specified, better screened or better priced when you look is the one to buy, and frame rates should play no part in the decision.

Quiet operation or guaranteed thermal behaviour is a hard requirement. We hold no noise or temperature measurements for this configuration, and on a machine sustaining 140W in a 16in chassis those are exactly the unknowns that would matter most. If you cannot tolerate the possibility of an audible machine under load, buy something for which that data exists.

The summary judgement

The Alienware m16 R2 is a machine that got the most important decision right and made a defensible compromise on the second most important one. The graphics card is configured at the top of its band and the measured results across ten titles show it behaving accordingly, with a wide, healthy spread that expresses the games rather than a bottleneck. The processor is a thin-and-light-derived part that costs it real performance in the minority of games that depend on the processor, and that cost is visible in exactly two of the ten published results.

For its generation and its era, that is a good gaming laptop with a clearly-defined blind spot, and the blind spot is described precisely enough in the benchmark data that you can check your own library against it before spending anything. What we cannot tell you is what it costs, how loud it gets, how hot it runs or how long it lasts unplugged. Those absences are worth weighing alongside everything above, because on a 140W machine they are not trivia.