Laptop vs. Desktop GeForce RTX 4090: How Much Do Nvidia’s Top GPUs Differ in Performance?

Nvidia’s GeForce RTX 40 Series graphics processors, built on the “Ada Lovelace” architecture, are extremely impressive (and, at the top end, expensive) performers on both desktops and laptops. The top dog in the stack, the GeForce RTX 4090, delivers chart-topping frame rates and is more efficient at advanced ray-traced lighting and image-sharpening DLSS than any before it. You can read both our laptop GeForce RTX 4090 testing analysis and our Nvidia GeForce RTX 4090 desktop graphics card review for context versus other GPUs in their respective spheres.

They share a name, but are they really anything like the same thing? How close are the desktop and laptop versions of the RTX 4090 GPU to one another, exactly? The gap between Nvidia’s full-fat and mobile GPUs of the same name has varied through the years, but across virtually all generations, it’s safe to say that desktop graphics cards always have had a distinct advantage. They are much larger, are actively cooled, are situated inside a desktop case with much more thermal headroom than a laptop chassis, and run at higher wattages. Laptop GPU performance and cooling technology have come a long way, but in the end, you can’t defy thermodynamics.

The Nvidia GeForce RTX 4090 Founders Edition graphics card
(Credit: Michael Justin Allen Sexton)

These clear, fundamental differences don’t change the fact that both are based on the same architecture, and sold under the RTX 4090 name, though. It’s easy for consumers to make assumptions about how performance translates between form factors or, for the less tech savvy, to simply overlook even the above factors and make assumptions based on the name alone.

We don’t usually put our laptop and desktop systems through the same exact benchmark tests, or pit them against each other in one chart—shoppers are generally looking to either buy a new laptop or upgrade their desktop, after all, not picking one over another.

The 2023 MSI Titan GT77 with an RTX 4090 GPU
(Credit: Joseph Maldonado)

To fairly judge how deeply two GPUs of the same name vary here in 2023, we’ve run the GeForce RTX 4090 Founders Edition GPU and an unconstrained-as-possible mobile GeForce RTX 4090 (inside the formidable MSI Titan GT77) through the same benchmarks and compared the results. Let’s set the stage, look at their technical differences, and then dive into the results.

What’s in a Name? Sizing Up the GeForce RTX 4090

I touched on it above, but the most obvious difference is that of size. Desktop GPUs, especially high-end models like the RTX 4090, are sizable physical objects before even accounting for the technology inside. The desktop chips are nestled inside big, thick cards that nowadays can be a foot long or more, while the mobile GPUs are thin chips that need to fit inside a laptop chassis, just like CPUs. The RTX 4090 Founders Edition, for example, measures 12 inches long and is considered a triple-width GPU—bigger hardware than any laptop could ever hope to contain.

A representational mockup of an Nvidia laptop GPU
(Credit: Nvidia)

In this particular case, another major differentiator appears: Despite the shared name, the mobile RTX 4090 actually runs on the same AD103 silicon as the desktop RTX 4080 graphics card, not the RTX 4090. The mobile and desktop GPUs share the RTX 4090 branding, as the top models in their respective stacks, but they don’t use the same silicon. This fact has a measurable effect on performance. And it means the laptop RTX 4090 has more in common, in terms of hard specs, with the desktop Nvidia GeForce RTX 4080 GPU than the graphics card that shares its name.

Here’s a look at those hard differences in the following table. For the desktop card, these specs are for the RTX 4090 Founders Edition. For the mobile RTX 4090, I’m using the MSI Titan GT77 as my test laptop; specs for this GPU in other laptops may differ based on wattage, which I’ll touch on below.

One big variable here is “board power,” as it relates to laptop implementations of Nvidia’s GPUs. A laptop maker can configure the mobile version of the RTX 4090 within a range of power delivery, depending on the laptop in question and its thermal limits. On the desktop side, a similar issue comes down to the individual graphics card makers. Nvidia makes Founders Edition (FE) graphics cards, which serve as the “reference,” or stock, version of the card. Partner manufacturers—Asus, Gigabyte, and MSI, to name a few—can take these reference designs and create their own versions of the card around Nvidia’s chips, often with more fans, clocked at a higher speed, equipped with heavier-duty, bulkier heatsinks, or some combination of these. The different cards have different power demands, performance, and pricing; we’ll just look at the Founders Edition card as a safe baseline.

The variances on the laptop side tend to be a bigger deal. You have no physical or standalone graphics card to purchase, since the GPU is always part-and-parcel of a laptop. But laptop makers can implement the chip with differing thermal constraints and power consumption levels. Those are tied largely to the size and cooling potential of a laptop’s body, so manufacturers deliver the same GPU with varying wattages in different laptops. In other words, an RTX 4090 laptop chip in one notebook isn’t necessarily given the same implementation as an RTX 4090 chip in another; never mind the difference between laptop and desktop!

For example, Asus chose to implement the RTX 4080 in its ROG Strix Scar 18 laptop at 150 watts (up to 175W, including boost), but another laptop maker may restrict it to 125W, limiting its power ceiling. Nvidia lists the range of non-boosted power for the RTX 4090 laptop chip as 80W to 150W. You’ll see this type of variation constantly with laptop GPUs, making looking at first-hand testing extremely important; the name of the GPU alone does not denote an exact performance expectation.

Testing the GeForce RTX 4090: A Desktop vs. Laptop RTX 4090 Face-Off

As mentioned, I’ll be using the Founders Edition card as the baseline desktop expectation for the RTX 4090, alongside the RTX 4090 deployed in the MSI Titan GT77. The desktop GPU was tested in our graphics-card testbed, which runs a stock Intel Core i9-12900K processor with a 240mm Corsair Hydro Series H100X liquid cooler on an Asus ROG Maximus Z690 Hero motherboard.

The Titan GT77 review unit I tested ran an RTX 4090 at 150W (up to 175W with boost), an Intel Core i9-13980HX CPU, and 64GB of memory. (Yes, the processor in the GPU testbed is a generation behind the laptop, as that test build was the standard at the time of RTX 4090 testing, and remains what we use to test desktop GPUs. The impact of any CPU throttling should be limited in this context.) The Titan GT77 is not the be-all and end-all of RTX 4090 mobile performance—some larger 18-inch systems exist and might push frame rates a touch further. But this is a widely representative high-end laptop that allows for the highest possible graphics wattage.

(Credit: Joseph Maldonado)

We had plenty else to say about the GeForce RTX 4090 Founders Edition, and the Lovelace architecture, so rather than rehash it here, I encourage you to read the details in that review if you’d like to know more.

With all of that established, the exact performance gap between the laptop and desktop variants of the “same” GPU has always varied within each of Nvidia’s platform generations. The launch of the RTX line and its advanced ray-traced lighting technology, in particular, created a new world of possible power requirements, and desktops have a lot more potential headroom than laptops. This made it even more difficult to compare them one-to-one, even if laptop GPUs had been drawing closer in performance prior to that time. (Back in the days of the GeForce GTX 10 Series, GPUs like the desktop GeForce GTX 1080 and its mobile GTX 1080 counterpart were actually getting pretty close in performance.)

Now, let’s get to the suite of benchmark tests. For these results, I included our test numbers for the desktop GeForce RTX 4080 Founders Edition card, as well. After all, if the mobile GeForce RTX 4090 is, at the core, based on the same silicon as the desktop RTX 4080, it’s just as interesting to see how closely these two perform. Seeing how the RTX 4090 name translates is the focus here, but the silicon being shared between the RTX 4080 desktop and RTX 4090 laptop chips will illuminate what’s lost in the different form factors.

High-Resolution and DLSS Testing

Much of the strength of the RTX 40 Series is in its DLSS efficiency, theoretically improving frame rates markedly at demanding high resolutions. This has proven the case with most of these GPUs, and the results here demonstrate that fact. Below are the results of two modern DLSS-equipped titles at maximum settings with DLSS toggled off and on, across multiple resolutions.

These games behave differently, and DLSS affects them differently, but you still see a pattern across these titles and GPUs. It’s fairly unsurprising to point out that, on average, they generally ranked as you’d expect—the desktop RTX 4090 was the fastest, then the RTX 4080, and finally the mobile RTX 4090—but the degrees of differences are interesting.

One exception is F1 22 with DLSS turned off—the mobile RTX 4090 was faster than the desktop RTX 4080 at all three test resolutions. It couldn’t come close to its desktop counterpart outside of 1080p, though, falling behind by 32 frames per second (fps) at 1440p and 17fps at 4K. Turning on DLSS only exacerbated this further—the mobile RTX 4090 posted 98fps at 4K, with the desktop version cruising to 147fps.

At 4K, 98fps is respectable and playable, but it’s still a sizable gap. The desktop RTX 4080 fell between the two at 4K, but the laptop RTX 4090 was actually a bit better at 1440p. This will vary by game. But, on this title, the mobile RTX 4090 if largely outperforms the desktop RTX 4080 while falling well behind its desktop sibling.

The caveat about game-by-game variance is important because you can see that this scenario immediately reverses in Guardians of the Galaxy. The mobile RTX 4090 can’t hang with even the RTX 4080 and is dozens of frames short of the desktop RTX 4090. Again, DLSS makes these titles smoothly playable at max settings with this GPU, and those frame rates probably sound decent to most mainstream players. But for enthusiasts shopping in the high-end tier, the laptop RTX 4090 performance in this game does not equate to the desktop expectations of hearing the name “RTX 4090.”

AAA Games Testing

This is a collection of titles that, while not all super new, represent a decent mix of genres and graphics fidelity in the modern market. We’ve tested them at three different resolutions, all at maximum settings.

DLSS effectiveness may be one of the main features of the RTX 40 Series, but with that stripped away, you can get a better sense of the raw performance here: The desktop-card RTX 4090 is much more powerful than its mobile counterpart. Red Dead Redemption 2 is the most visually demanding title here, and the difference at 4K, in particular, is stark.

The desktop RTX 4090 is much more powerful than the RTX 4080, too, but the mobile RTX 4090’s deficits throughout here make it look like a GPU much further down the stack. In practice, that’s what the mobile GPU really is in terms of performance. The numbers illustrate that the RTX 4090 name does not ensure a certain level of power everywhere it appears.

In some instances on these game tests, the mobile RTX 4090 hangs relatively close to the desktop card, but generally speaking, the frame rates fall into wholly separate tiers. You’re looking at gaps of as much as 83fps at 4K in some of these games, and as much as 67fps in some instances at 1440p. Not all games see such a big swing, but the overall takeaway remains the same. The performance delta isn’t nearly as large as with the RTX 4080, but the inherent desktop card advantages remain there, too.

To underscore the raw power difference, I also ran the RTX 4090 mobile through Futuremark’s synthetic 3DMark Time Spy Extreme test…

It scored 10,622 points, to the RTX 4090 FE’s 19,264 points and the RTX 4080 FE’s 17,672 points. No contest there.

Verdict: There’s Power in a Name, But Don’t Mistake the Laptop RTX 4090 for the Desktop Card

Ultimately, that’s the reality of these results. Regardless of the GPU names, mobile GPUs are always at a power and thermal disadvantage, and the performance delta you see here is not too surprising given the differences I outlined at the top. Hopefully, this illuminates exactly what those variations in frame rates can look like in practice. The GeForce RTX 4090 mobile does hew more closely to the graphics card it shares a base chip with (the RTX 4080) than it does the graphics card it shares a name with, but even then, the mobile chip can’t keep up.

That’s not to say calling this mobile GPU an RTX 4090 is intentionally misleading—it represents the top-of-stack GeForce RTX 40 Series option for laptops specifically. If you want maximum performance (and not necessarily value, as we’ve seen mobile RTX 4080 systems perform nearly as well), this is the GPU to get. But it is something of a misnomer, as it’s more a name for branding consistency than it is a hardware or performance guarantee. Laptops with the RTX 4090 GPU are meant for the deep-pocketed enthusiast, and will absolutely deliver smooth frame rates at the highest settings—the chip just won’t meet the raw performance bar set by its much more powerful desktop counterpart.