All right, with me is Asus’s updated Zephorus G16 for 2025. If you haven’t heard of this laptop, it is really the laptop that you should be considering if you’re looking for the balance of performance, premium features, a large screen, and you want maximum portability. This combination is really hard to do well in a laptop. It’s kind of the best of everything. In fact, speaking personally, this is the category that got me into this game. I was looking for a laptop like this and I couldn’t find anything, at least not without massive compromises. Well, last year the Zephris G16 was fully redesigned into the chassis you see right here, but it got off to a rough start. Both I and other reviewers found issues with it. Anyway, after months of updates and the swap to AMD’s Ryzen processors, things got a lot better and the G16, it ended up being the best portable gaming laptop of 2024. Well, this year we have a switch back to Intel with their new AiraLake H processors and we get Nvidia’s new Blackwell GPUs inside. When it comes to the GPU, generation on generation, we see the usual disappointing 15% or so bump in performance. Unfortunately, this is the norm for these 50 series GPUs. However, with all the changes combined, this year’s G16 is actually much improved from last year’s. It feels noticeably cooler to the touch, and the 5070 Ti version has more VRAM than the older 4070, which is better for driving that 2560 x600 resolution display. Plus, this year’s G16 also has noticeably better battery life. But before you run out and buy one, I’d avoid buying an RTX 5080 or 5090 config. You’ll just be throwing money down the toilet. The GPU’s wattage in this laptop is capped way below the 175 watts that others run this GPU at. So, the Zephris G16 significantly underperforms competitors with a 5080 or 5090. And with that said, here’s how we came to these conclusions. Thanks, Josh. Our version of the Zephris G16 has Intel’s AeroLake 285H CPU, the RTX 5080, and 32 gigs of memory. Let’s start by looking at synthetic benchmarks like 3D Mark Time Spy. Here we are seeing a pretty minor improvement over last gen, just 11%. Looking at the CPU component of this test, Aerrowake H only outperforms Meteor by 7%. Keep in mind during gaming, CPU’s wattage is limited to allow the GPU to perform better. We’ll talk more about this as we get into the video. The 5080 GPU only does about 12% better for graphics than the 4080, and this year’s G16 feeds its GPU 5 watts more than last year’s. So, it’s a very disappointing upgrade. Let’s now compare the results of the new G16 to the Blade 16. The Blades Ryzen 9365 CPU does worse than the AeroLake 285H CPU, but its higher wattage 5080 performs so much better that it carries it to the win in graphics and overall score. Josh talked a lot about this in his review of the Blade 16 that the pairing of the AMD Ryzen 9365 processor with the 5080 is just bad. Video linked below. The same patterns can be seen on Wildlife Extreme, which we primarily use so that we can compare with MacBooks. Here you can see that the 5080 card is clearly capable of beating the 40 core GPU in Apple’s M4 Max chip because it does so in other gaming laptops. However, ASUS not feeding their RTX 5080 full wattage is hurting it a lot here. These graphs all pretty much came out the same. So here’s this pattern seen again in Port Royal, a rate tracing benchmark, and then finally in Steel Nomad, which is a 4K gaming benchmark. At this higher resolution, 41 FPS is certainly more playable than the 35 or 26 of the prior year’s versions. However, many people would not find that to be super enjoyable. The laptop’s internal display is 2560 x600, not 4K. So, here all of our gaming results run at their native resolution. By the way, watch for the change in the order of our graphs. Unfortunately, Excel is limited on how we can order the laptops. The more powerful laptops are going to be at the top in just these specific FPS graphs. We are planning to move to a more advanced graphing software in the future. In our first game, Fors of Horizon 5, the uplift is slight, about a 10% increase in FPS from last year’s 4080 and even less than that for 1% lows. You can also see here how much better the higher wattage 5080 does in the Razer Blade 16, which has a similarly thin chassis. But even the Blade doesn’t perform as well as bigger, chunkier laptops with an RTX 5080 like the Legion or the Omen. You’re not getting anywhere near the full performance of a 5080 in a thinner chassis, even when it’s higher wattage. This issue remains the case across all our gaming tests, like for instance Monster Hunter Wilds. Here, the gap between the Blade and the G16 narrows a lot, and the lift from the older 4080 G16 is about 11% instead of 10%. We also see a bigger improvement in 1% lows, about 30%. Interestingly enough, the older G16 with the 4070 does better than the 4080 here. We believe this is due to the CPU’s power efficiency. The AMD Zen 5 CPU in our G16 with a 4070 is more efficient than the older Intel Meteor CPU in our G16 with a 4080. The same goes for this year’s AeroLake H CPU. It is more power efficient than Intel’s older Meteor CPU. We believe that more power efficient CPUs perform better at lower wattage during challenging parts of the game. That’s our hypothesis at least. And we’ve seen these increases in 1% lows play out in most 2025 gaming laptops with Intel’s AeroLake H or HX chips when compared to the older Intel chips that they replaced. We’ll talk more about efficiency later in our CPU performance section. But stepping back and looking at these results as a whole, none of these thinner gaming laptops do a great job on this benchmark, only getting around 60 FPS on average, which is really the bare minimum. To be fair, this is the most demanding game on our list. Next, we’ll look at Final Fantasy Dawn Trail. This is a game where many of the 50 series don’t see improvements over their 40 series counterparts. Surprisingly, when it comes to the G16, we do see a small increase in FPS and 1% lows over the prior model. It also does better than the Blade here. Please note, this game is a bit more CPU dependent. We’ve observed lower GPU wattage than what you’d expect in most of these 50 series laptops during Dawn Trail. Other than the G16, which still feeds its CPU the same wattage as it does in other games. A new Nvidia driver came out during the time we were making this video. We did retest with it, but this issue still doesn’t seem to be resolved. Lastly, Cyberpunk. As you can see, it’s a similar story. This year’s G16 is a small amount better than last year’s. Now, in this test, the uplift from 40 series to 50 series is about 19% when you have frame gen off. The Blade performs better in average FPS as it’s a higher wattage 5080. It performs worse with 1% lows. This may be due to the less powerful Ryzen 9365 processor in our version of the Blade versus the Ultra 9285H in this year’s G16. Once again, you can see a big jump in gaming performance to larger gaming laptops with these same graphics. Now, let’s take a look at gaming performance in Cyberpunk when we turn on DLSS frame generation. For the record, when we test DLSS frame gen, we do so on 2x mode. You would likely see better results from the 50 series in 4x mode, but that introduces additional latency issues. We believe most gamers won’t want to run on that mode, so we aren’t testing it here. With frame gen on, you do see a performance boost, of course, by 57%. The 5080 G16 is now 17% higher than the 4080 G16. The boost from using Frame Gen is actually slightly greater for the 40 series than the 50 series, 59% versus that same 57%. In creative tasks like Premiere Pro, we now see a bigger 30% performance boost moving from the 4080 to 5080 G16. You do get an additional H.265 10- bit encoding and decoding in these new 50 series GPUs, though. Some video editors will value this. In this test specifically, the G16 now matches the Blade. This is likely due to it having an Intel CPU. Premiere Pro tends to work better off of Intel than AMD. By the way, we are planning to include Da Vinci Resolve in our benchmarks as soon as it works on this new hardware. they still get stuck at the neural optimization step which skews the results. Now, let’s take a look at CPU performance. If you’re doing CPU intensive tasks outside of gaming, here’s how the laptop stacks up to last year’s. In Geekbench, we see about a 10% jump in single core and 18% jump in multi-core. Here you can see that the Aerolake HX chip in larger gaming laptops is much faster, as is Apple’s highest end M4 Max chip. These chips both blow the other laptops out of the water, at least for this test. Keep in mind, Geekbench tests a variety of common performance tasks and can benefit from various hardware optimizations, including memory speed. With that said, let’s move over to Cinebench, which tests the processor when it’s pushed to its max. Here, we can see the gap narrow between Intel’s AeroLake H and AMD Zen 5 and last year’s G16, at least in multi-core. In single core, Intel has the win. Compared to Aerolake HX found in much larger gaming laptops, they perform nearly twice as fast as these thinner gaming ones. Unfortunately, these thinner gaming laptops are thermally limited and can’t support the power delivery required by one of those HX chips. Larger gaming laptops with them feed their CPUs around twice as much power as what the G16 feeds its CPU. Between Intel’s Aerolake H and AMD’s Zen 5 CPUs, they are very close in power efficiency. The Intel chip pulls a little bit more power for similar performance. This is evident when you look at our scatter plot, which shows Cinebench multi-core performance at different power levels. Here you can see that this new Aerolake H CPU slots in between the Ryzen 9 HX370 and the Ryzen 9 365. It also slightly beats out the older Intel 185H chip. You can clearly see that gaming laptops with an Intel Aerolake HX chip have a huge advantage, very noticeable at the lower power levels of 50 to 80 watts that manufacturers limit them to during gaming. You may also notice we are missing some dots on the 285H here. This is because I wasn’t able to consistently set up manual wattage limits on the G16. We tried Armory Crate and Ghelper both separately, and multiple restarts later, it simply wasn’t letting us change the power limits at the level of granularity that we could on prior G16s. We are hopeful this is just another example of 50 series software or driver bug that will be resolved soon. Now, the reason we care so much about efficiency is that it determines a lot of other things about using the laptop. Efficiency tends to be the most important factor for keeping the components as well as the laptop itself cool during performance loads, which this new G16 does quite well. Its average CPU temp during Cinebench stays about 10° cooler than last year’s, even though it maxes around the same. You might also notice on this graph that the AMD version of the G16 gets quite hot as well, even though the Zen 5 chip is normally pretty efficient. Our theory is that this may be due to the lesser cooling solution in the lower-end versions of that laptop. When you buy the G16 with a 4080 or 5070Ti and above, you’re getting a vapor chamber cooler rather than the three fan cooling solution of the 4070 and 5070 and below. This improved cooling solution and more efficient CPU leads to a much cooler underside in this year’s model. The fan noise is even a tiny bit quieter, although that could just be a measurement error at only 1 decel. For gaming tests like Time Spy, the G16 with a 4080 and Meteor Lake got up to 50° C on the underside, whereas this new one stays at a super comfortable 36° on both the keyboard and underside. The only other laptop with a similarly impressive cooling solution is the Legion Pro 7i, which is much bulkier and heavier. This is another rare place where the G16 beats out the Blade, which gets warmer during load. This improved cooling seems to be a major benefit, if not the biggest benefit, of getting the new G16 over the old one or the Blade. When we test these laptops, we do test them in their different performance modes. In its quiet mode, it drops its power down to just 20 watts. This is similar to a lower powered Lunar Lake laptop and performs as such. Luckily, you aren’t losing any performance between the standard performance mode and turbo. So, you can stick to performance for demanding CPU tasks for lesser fan noise. To see how the performance modes differ in gaming, we ran time spy for a few as well. Here, we see the GPU throttle quite a bit between performance and turbo. So, that’s something to look out for when you decide which performance mode to game on. Now, we did try Ghelper because you guys recommended it. It’s a third party tuning application that’s meant to replace ASUS’s own Armory Crate. In our gaming testing, manual mode and Armory Crate and Ghelper did not have much of a positive effect, if one at all. That’s why all of our results so far are using turbo mode. Another improvement of the new over the old G16 is battery life. For light use, it lasted an hour longer than last year’s model. Over 10 hours for a gaming laptop is extremely impressive. But this is in Armory Crates eco mode, which disables the DGPU and on quiet mode, which as I showed you significantly throttles its performance. As a worst case battery life test, we ran Cinebench on a loop until the laptop died. It took only an hour to run out of battery. Unfortunately, this is pretty comparable to other similar laptops. The MacBook looks bad here, but scored by far the best in Cinebench during this test. So, in a real world performance scenario, it can complete its task faster and go back to a lower powered state. We don’t tend to recommend heavy performance tasks like gaming or rendering a video on battery, though, and this is exactly why. Now that we’ve gone through all of the big changes to the G16 with this new version and its components, let’s talk about what hasn’t changed much or at all. Starting with the upgradability, you can still replace the Wi-Fi 7 card, battery, SSD, and add a second SSD. Memory is still soldered. Now, let’s take a look at the chassis. Since it hasn’t changed at all this year from last year’s design, we’ll go over how it compares to ones in the same class, such as MacBook Pro 16 and the Razer Blade 16. Firstly, when it comes to size and thickness, these three are nearly identical. I expected the Blade to be the thinnest, but it’s actually not. The G16’s palm rest goes to about the same height as the Blade, and the screen is very slightly thinner on the G16. It doesn’t make the G16 screen less sturdy, though. It’s still just as strong. The build of the MacBook feels the sturdiest. Deck flex continues to be minimal on the G16. It still feels very well built. The gray version we have does pick up fingerprints a bit like the MacBook with both of their Mac coatings, but not as badly as the black blade. As far as weight goes, these are all within half a pound of each other, but the G16 does come out as the lightest by a smidge. You can also open the screen with one hand. When you do, you’ll be greeted by a 2560 x600 OLED panel that we measured at around 440 nits of brightness. It also has a 240 Hz refresh rate and is not a touchcreen. This panel seems to be the same as last year’s except it does get a little bit brighter, which is nice to see. That being said, it’s still a glossy display, and at this brightness, it’s not enough to truly combat reflections in a very bright environment like outdoors. Compared to the Blade, this screen is a bit brighter, but the MacBook Pro does have a higher resolution and brighter screen than both of them. Back on the G16, we did not see any noticeable screen door effect or detect any PWM flickering, which is good. This display also supports G-Sync, which is meant to help with screen tearing, and advanced Optimus, which helps you switch GPU modes. Next, we’ll look at the keyboard, which I’d call fine enough. It has a standard layout and decent key travel, but it feels a little mushy. When you compare it to either the Blade’s new keyboard or the MacBook, it simply isn’t as clicky or comfortable to type on. It’s not uncomfortable, though, just fine. The RGB backlighting is a nice gamey touch, but doesn’t look as clean as the Blades. The trackpad is similar. It’s basic and it isn’t loud, but the click is a little deeper than I like to see on a mechanical trackpad. It doesn’t feel as soft as the Blades one. The MacBook Pro wins here with its very accurate haptic trackpad. The ports are the exact same as last year, which means a fast SD card reader. Yay, but no Thunderbolt 5. We’ll put the full list on screen for you. Now, the 1080p webcam on the G16 is the same as last year’s, but it’s pretty disappointing for this price point. The colors are washed out and inaccurate, and it doesn’t sound great either. Speakers sound really nice on this laptop, but weirdly a little less deep than last year’s. Take a [Music] [Applause] listen. When we tested Linux, the only thing that didn’t work was the display’s brightness. Everything else seemed to function just fine. Wi-Fi, Bluetooth, camera, speakers, you guys get the point. All right, let’s wrap. Firstly, big thanks to Sierra for doing the crazy amounts of testing that you just saw. We recently changed a number of our tests to make them more relevant for you guys. we need to improve is in battery testing. We’re kind of giving you either ends of the range. The worst case battery results you could possibly get and then the best case. Anyway, here’s our advice when it comes to the Zephris G16 for 2025. Look, last year we thought it was the best lightweight gaming laptop that you could buy, and we think that this year’s model has improved, but only if you get the RTX 5070Ti configuration. Above that, the GPU is just power limited and underperforms. Below that, the 5070 version, it only has 8 gig of VRAM, which we don’t feel is sufficient for driving games on a high resolution display like this laptop has. Unfortunately though, we didn’t get the 5070Ti model in. So, the best we can do is extrapolate how it will perform based on our RTX 570 Ti G14 and the Stricks G16. Their time score ranges from 14,000 to 17,000. The G14 has a similar powered processor and only feeds its 5070 Ti, 5 watts less power than the 125 W of the Zephris G16. Therefore, we feel the 5070Ti version of the G16 should get around a 15,000 score in Time Spy. That’s if we’re being generous. So, this means that the 5070 Ti version is dropping around 1,500 points in Time Spy from the 5080 version that we tested today, i.e. 9%. Given that the 5070 Ti version costs around $700 less, we feel that trade-off is definitely worth it. So, here’s where we land. If you want the best gaming performance in a portable 16-in laptop, the Razer Blade 16 is the way to go right now. That being said, that laptop is a bit more expensive and certainly has its own set of unique issues. If you want to see me just rant about them, then make sure to check out the video which I’ve linked below. If you are okay with 5070 Ti level performance and you can wait, I’d recommend holding off for reviews of Asus’s new Predator 16S. That’s their entry into the slim gaming market. Aser is known for making excellent gaming laptops and it has a much more powerful CPU in it. Now, that may make it run hotter. Look, we expect to have a review out of that laptop in around 2 months or so. If on the other hand, you do want to buy a really good portable gaming laptop right now, we highly recommend this Zephris G16 with a 5070Ti. It makes solid improvements from last year’s model, particularly when we compare it to the similarly priced older version with the 4080. The new model gives you similar GPU performance, but has a slightly better CPU. It is noticeably cooler to the touch and has better battery life. Finally, if you aren’t We have a ton of videos coming. Till next time, go do something awesome with your day and I will catch you later.