Earlier this month, Nvidia announced it would be shipping mobile versions of its GeForce RTX 2070 Super and RTX 2080 Super GPUs on laptops. Now, we’re able to bring you some early first-hand testing to see how they stack up against existing GeForce laptop GPUs.
In the time since the announcement, we were able to get hold of and start testing two laptops that carried the GeForce RTX 2070 Super and the RTX 2080 Super (both of them in dialed-back Max-Q configurations). The former is in the Gigabyte Aero 15 OLED XB, while the latter is installed in the MSI GS66 Stealth, both part of a new 2020 wave of gaming laptops that boast Nvidia GeForce RTX Super GPUs and Intel’s 10th Generation Core H-Series processors.
We were able to do this testing with early versions of the soon-to-be available public drivers, and of course only have these two samples to go on so far. As a result, these early tests below may be representative of what these components can do, but bear in mind: This is just two laptops, an early sampling, with first drivers and lots of variables between and among them and earlier laptops we’ll use for comparison.
An ‘RTX Super’ Refresher
For background, these top-end graphics chips will, in theory, push the higher end of Nvidia’s laptop-GPU offerings even further for laptops. On the desktop, the GeForce RTX Super series delivered better performance than the original graphics cards at a lower price point. In practice, Super-izing the GeForce RTX 2070, for example, turned it nearly into an RTX 2080, at a lower price than the original stock model. You can see our reviews of the GeForce RTX 2070 Super and GeForce RTX 2080 Super desktop cards for many more details and lots of testing.
Like with the desktop cards, the laptop versions of these GPUs should give hardcore gamers even higher frame rates in AAA games, make it more likely you can use Nvidia’s advanced ray-tracing technology without tanking frame rates, and serve as a boon for high-refresh-rate gaming (as long as the laptop’s display panel allows it). You can see the new Super GPUs in relation to the rest of the Nvidia laptop-GPU lineup here…
As I wrote in my original analysis, until this launch, the original GeForce RTX 2070 and RTX 2080 were the premium GPU options in Nvidia’s laptop-GPU stack, so these new upticked Super versions should push performance further in elite, enthusiast-grade gaming laptops. You will not see them in value-focused laptops, though the existence of new high-end options does have a knock-on price effect at the entry level, which will be lower going forward. For example, Nvidia notes that shoppers should see laptops with the GeForce RTX 2060 starting at $999, a new low for RTX-based gaming machines.
The Gigabyte Aero 15 OLED XB and the MSI GS66 Stealth are two such premium laptops, with slim designs and an array of advanced features in addition to the powerful GPUs. You can read our full review of the Gigabyte Aero 15 OLED XB today, while the MSI GS66 Stealth arrived scant hours ahead of this article, so a full review will be forthcoming soon. (We got it on the bench for a few initial tests, though.)
One key twist on this testing is that these laptops are using, as I mentioned earlier, the Max-Q versions of the RTX 2070 Super and RTX 2080 Super. If you aren’t familiar with the term “Max-Q,” you can read more about it here. But the short explanation? It’s Nvidia’s approach to getting its more powerful GPUs into slim and light laptops by down-tuning their power potential to restrict thermal output.
Since the Gigabyte and MSI models here are more-portable 15-inch laptops (as opposed to 17-inch models), going with Max-Q tuning was the obvious choice to keep them relatively thin, so these versions of the GPUs presumably won’t be as powerful as the unrestrained RTX 2070 Super and RTX 2080 Super. There’s a good chance we’ll end up reviewing more Max-Q laptops than non-Max-Q ones, given the modern focus on being thin and light, so it’s important to test both kinds.
We’ll have to wait until review units of larger laptops show up bearing these “base” non-Max-Q GPUs to measure their raw power. But considering how early we are in the stages of this launch, we’re content to have two RTX Super laptops to analyze. Still, between Max-Q and Super-izing these GPUs, there are a lot of factors affecting the results, so keep that in mind.
Points of Comparison, and a Growing List of Variables
For testing comparisons, I gathered a batch of past laptops we’ve reviewed with the relevant GPUs. Given that we’re testing the GeForce RTX 2070 Super (Max-Q) and the RTX 2080 Super (Max-Q), there are a few obvious inclusions as points of reference. Below is a cheat sheet of their names and specs, with the two new RTX Super laptops up front and the rest in (theoretical) ascending order of GPU power.
Obviously, it was imperative to include machines with the non-Super RTX 2070 and RTX 2080. Since both of the new laptops employ Max-Q, I went with Max-Q versions of the originals as well, in the form of the Alienware m15 and the Razer Blade 15 Advanced Model, both of these 2019 models with OLED screens. This comparison allows us to get as close as we can to seeing what difference the “Super” aspect of the GPU makes. I also included a non-Max-Q RTX 2070 laptop from Acer, to see if the Super version of the GPU adds more power than Max-Q takes away. None of these is a flawless comparison, but it should give a rough idea of the scale of differences (or lack thereof) among them.
Sadly, one variable we can’t eliminate is the processor differences. These are the only two RTX Super laptops we have to test for now, and they carry different CPUs. While they are both part of Intel’s new 10th Generation H-Series, they are fairly different. The Gigabyte’s “Comet Lake-H” Core i7-10875H is clearly the more muscular of the two, boasting eight cores and 16 threads compared to the six cores and 12 threads of the MSI machine’s Core i7-10750H. This means it has more throughput for strenuous tasks, and while that is generally more relevant to media-processing workloads, it can still affect gaming performance. This should also add to the performance gains compared to the laptops using older Intel chips.
Just to make this comparison even more byzantine, you’ll notice that the better CPU of the two happens to be paired with the lesser GPU of the two, further fogging up clear results. Again, we’ll do our best to draw conclusions from these results, but there are so many mitigating factors to take into account, that it’s harder than ever to draw one-to-one comparisons between new laptop components. In an ideal world, we’d be doing a head-to-head faceoff between the two GPUs in laptops with the same CPU and RAM capacity (even better, in the same laptop!), but this is the early hand we were dealt.
If anything, you’ll see that this will only go to strengthen the value of full reviews of gaming laptops. You simply can’t draw conclusions from lists of specs alone. Indeed, as more and more variables are added (different combinations of CPU and GPU, Max-Q or not Max-Q, RTX Super or not, varying chassis and thermal designs), it’s harder than ever to “unbake” the results from the cake that is the laptop’s ingredients. Objective test results are all we can go on.
With those many factors in mind, on to our first Super results…
Now Testing: GeForce RTX Super Laptop Analysis
I’ll focus on the Gigabyte Aero 15 OLED XB and its RTX 2070 Super (Max-Q) first. Both of the new GPUs are in these charts, but it’s easier to digest the results by zeroing in on one at a time.
On paper, it should be superior to at least the Alienware’s RTX 2070 (Max-Q), and depending on whether the Super up-tuning or the Max-Q down-tuning has a greater impact, it may perform better than the Helios 700’s RTX 2070. If you asked me before testing, I may not have expected it to post better numbers than even a Max-Q RTX 2080, but let’s see what the results had to say.
The GeForce RTX 2070 Super (Max-Q) in Gigabyte’s Aero 15 OLED XB
First up is 3DMark, which measures relative graphics muscle by rendering sequences of highly detailed, gaming-style 3D graphics that emphasize particles and lighting. We run two different 3DMark subtests, Sky Diver and Fire Strike, which are suited to different types of systems. Both are DirectX 11 benchmarks, but Sky Diver is more suited to midrange PCs while Fire Strike is more demanding and made for high-end PCs to strut their stuff. The results are proprietary scores.
Next up is another synthetic graphics test, this time from Unigine Corp. Like 3DMark, the Superposition test renders and pans through a detailed 3D scene and measures how the system copes. In this case, it’s rendered in the eponymous Unigine engine, offering a different 3D workload scenario for a second opinion on each laptop’s graphical prowess.
These results are great news for the RTX 2070 Super. On all four versions of these benchmarks, the Aero 15 performed better than all of the non-Super laptops, period. It also outperformed the MSI GS66 Stealth and its 2080 Super Max-Q, but we’ll get to that seeming anomaly later. For these higher-end GPUs, 3DMark Fire Strike and the 1080p setting for the Superposition trial are the tests to focus on as the more strenuous benchmarks, and it bested its predecessors.
This includes even the RTX 2080 (Max-Q) in the Blade 15, showing that the added power of the Super series outweighed the performance cap from the Max-Q design, at least in this case. (Of course, it could also be down to driver improvements made since the test of the Razer, thermal differences, or something else.) Plus, a quick scan of our results database shows that it beats all other GeForce RTX 2080 Max-Q laptops we’ve tested in these benchmarks, as well, not just the Blade 15.
The Aero didn’t top everything we’ve tested. Larger laptops bearing the regular, non-Max-Q GeForce RTX 2080 posted higher scores. The Super tweaking was not enough to close that gap when held back by Max-Q, it would seem. I’m definitely curious how close an RTX 2070 Super without Max-Q would come to an RTX 2080, though, and can’t wait to test that when we get a chance. It is also worth remembering that the Aero 15 is benefiting from that beastly new eight-core processor, helping its performance.
Next up are the real-world game tests. The synthetic tests above are helpful for measuring general 3D aptitude, but it’s hard to beat full retail video games for judging gaming performance. Far Cry 5 and Rise of the Tomb Raider are both modern, high-fidelity titles with built-in benchmarks that illustrate how a system handles real-world gameplay at various settings. We run them at 1080p resolution at the games’ medium and best image-quality settings (Normal and Ultra for Far Cry 5 under DirectX 11, Medium and Very High for Rise of the Tomb Raider under DirectX 12).
Performance will always vary markedly from game to game, but even so, it’s more positive viewing for the RTX 2070 Super (Max-Q). On Far Cry 5, it repeats its top-of-the-heap performance on both graphics presets, and by a comfortable margin. For mobile gamers, this is very promising data if it’s representative of a trend. It would indicate that, slowly but surely, the more potent GPUs are hitting a new tier of frame rates suited for high-refresh screens, even in big-budget games in slim laptops. You won’t hit as many frames per second as you will in competitive multiplayer titles, but it’s still closer to the common 120fps or 144fps targets than it was before, and well above the 60fps acceptability floor.
The results weren’t as strong on Rise of the Tomb Raider, actually ending up at the bottom of the pile, so something about the design of this GPU or the driver may not play nicely with this game. Again, variance by title is pretty common, even if it’s unfortunate that you may have to expect lower frame rates on some games. Hopefully more games trend in the direction of the Far Cry 5 performance for this GPU. Rise of the Tomb Raider may be a one-off for this laptop, as well, because it should at least be pushing higher frame rates than the non-Super RTX 2070 Max-Q, and it did not here.
All in all, though, it’s a strong showing for this laptop and the RTX 2070 Super (Max-Q). It may have done even better than expected because of the healthy eight-core assist from the processor. Before we get too deep into drawing any conclusions, though, we’ll move on to its stablemate.
The GeForce RTX 2080 Super (Max-Q) in MSI’s GS66 Stealth
This one is, unfortunately, less cut and dry. With no need to recap what each benchmark is measuring, I’ll start right in with the synthetic results. These are the same charts as above, posted again for convenience.
The GS66 Stealth and its RTX 2080 Super (Max-Q) bested all of the older laptops on these tests, with the exception of the Acer Predator Helios 700 on Fire Strike. It’s good, but not great viewing for a GPU that should be upping the ante, and makes us wonder if the machine is hampered by its Max-Q design, even with its Super uptick. It is still more potent than the non-Super RTX 2080 Max-Q here, though by a very slim margin on Superposition.
I mentioned previously that I’d get around to the fact that the Gigabyte’s RTX 2070 Super (Max-Q) GPU outperformed the MSI across the board despite theoretically being lower in the GPU hierarchy. At first, this is one of the most eyebrow-raising aspects of these results, though there’s one obvious X-factor here. As mentioned earlier, the Gigabyte laptop’s superior processor, with more cores and threads, might be paying real-world frame-rate benefits.
The MSI’s Core i7-10750H could be a relative bottleneck, rather than it being the fault of the RTX 2080 Super (Max-Q). Logic tells us the RTX 2080 Super should be the better of the two GPUs and, when the processor with a greater core and thread count is combined with a higher RAM total, it’s eking out better performance than the superior GPU can with less RAM and a lesser CPU. As such, I wouldn’t be too harsh on the RTX 2080 Super (Max-Q), especially in light of our real-world game tests. Here are those charts as a refresher…
The gap between the new RTX 2080 Super (Max-Q) and its non-Super counterpart in the Razer Blade 15 opened up much more here than on the synthetic tests. This may be closer to what you’d expect from the top-of-the-line Max-Q GPU, though the Gigabyte’s advantage held true on Far Cry 5 as well. Its unusually low Rise of the Tomb Raider numbers, though, mean the MSI GS66 Stealth was the better of the two there. The Max-Q aspect does ultimately hold it back, though, since the Helios 700 and its full-fat RTX 2070 was the best of the bunch on that game. Again, I look forward to when we can test the non-Max-Q version of these Super chips.
No Simple Answers, But Super Brings More Power
We can take a couple of major lessons away from all of this quick-take testing. Largely, the RTX Super GPUs are in fact improvements on the existing base versions. This isn’t 100 percent evident in all tests, and many outside factors affect the results on a per-test basis, but the added performance potential is clear. These are the Max-Q versions of these Super GPUs, after all, and it is not difficult to imagine the full-strength mobile RTX 2070 Super and RTX 2080 Super chips easily topping all of these charts.
The other big takeaway is the importance of testing these laptops ourselves. We as experts (and many of you readers, as enthusiasts) can speculate and make plenty of educated guesses about how these laptops will perform. But it’s almost impossible to say which will come out on top without rigorous, controlled testing. Nvidia has added four more variables to the mix with the standard and Max-Q versions of these two Super GPUs, on top of the existing hierarchy.
When you compound those with multiple processor options (we haven’t even brought AMD and its new Renoir CPUs into the mix here), with differing core counts, thread counts, and clock speeds, you end up with myriad permutations that will interact with one another differently. Ultimately, this results in already thinly sliced performance tiers being sliced even more thinly, and the true power differences obscured. Sure, all that variety means a laptop market with more options for a wider array of budgets and needs. But it also makes purchase decisions more confusing.
We’re convinced that this ever-growing complexity makes our in-depth testing and reviews even more valuable going forward. The gaming laptop market is getting increasingly complicated, and any one component is harder than ever to “unbake” from the whole. The bottom line here is that if you’re a hardcore gamer shopping at the high end of the gaming-laptop market, the new Nvidia GeForce RTX Super GPUs are definitely promising on a power basis. You may end up quibbling about which is the better value, which we’ll be able to tell you once more laptops become available to test and review. But based on this initial flurry of tests, greater performance is headed your way.