caption

Laptop vs Desktop (RTX 3060) – Closer Than You Think!


The difference between Nvidia’s RTX 3060 desktop and laptop is way smaller than you think it is! The laptop actually wins in some games. I’ve compared both in 17 different games at 1080p and 1440p to find out all the differences. These are the differences in specs between the 3060 mobile and desktop GPUs. Interestingly the 3060 for laptops has 7% more CUDA cores than the desktop version, but the desktop card has double the VRAM capacity with 12 gigs.

The desktop card also has a higher power limit, and can clock higher as a result. The laptop 3060 has to have lower power limits because more power generally equals more heat, and thermals are obviously a much bigger consideration in a smaller laptop form factor compared to a desktop PC. Seriously, just look at the size differences, there’s way less space for cooling in the laptop.

To do this testing I’m using the Lenovo Legion 5 for the laptop, it’s got an 8 core Ryzen 7 5800H processor and 16 gigs of x8 memory in dual channel. I’ve chosen this model because it both runs the 3060 at its maximum power limit and it’s also got a MUX switch so we can disable optimus – which is something that desktop PC’s aren’t subject to.

Now technically the Nvidia spec sheet does allow laptop 3060s to go all the way down to 60 watts, so definitely expect lower results there compared to what I’ll be showing with this best case 3060 laptop. The desktop PC on the other hand seems kind of overkill comparatively, with 12 core Ryzen 9 5900X processor as unfortunately I couldn’t get the 8 core, and with 32 gigs of DDR4-3200 CL14 memory it’s got both faster CPU and RAM, but before you run into the comments to type “unfair comparison, desktop is way more powerful, unsubscribe!” I’ll give you a quick little spoiler and note that the laptop can still beat the desktop.

Alright let’s get straight into the 17 game comparison followed by things like pricing & availability, content creator workloads and power draw afterwards. Let’s start out with Red Dead Redemption 2, which was tested with the games benchmark.

I’ve got the 1080p results below and the 1440p results above. At 1080p the desktop 3060 was only 7% faster than the laptop 3060, then at 1440p the desktop card was 11% higher, or just 7 FPS, but the laptop was still capable of more than 60 FPS at high settings here. They both get an extra boost if we enable DLSS with quality mode, the desktop card now has a larger 17% boost to average FPS at 1080p, so it appears that the gap increases with DLSS on.

Cyberpunk 2077 was tested in the same part of the game on both machines. At 1440p the laptop is actually slightly ahead, I mean it’s like one and a half frames, but hey winning’s winning, even if realistically this is probably margin of error stuff.

At the lower 1080p resolution the desktop was just 3% faster in terms of average FPS, however there’s a larger 13% boost to the 1% low.

It’s a similar sort of deal if we enable DLSS, the laptop 3060 was still slightly ahead at 1440p, at least in terms of average FPS, and again at 1080p the desktop was winning, but the 1% low difference increases to a 27% improvement now, so a more stable experience with the desktop system. Call of Duty Warzone was tested with all settings maxed out. The desktop graphics card was back in front now, but only by 4% or so, or about 3 FPS at 1440p, or 7% at 1080p, so although a win, it’s not that big of a difference, I wonder if you’d actually notice the difference in a blind test? This is another game that has DLSS support, and both systems are able to get a further speed boost by enabling this simple setting.

Control is a fairly GPU heavy game, even at 1080p, though the desktop graphics card was only reaching 4% higher average FPS here. The gap increases much more significantly at the higher 1440p resolution, with the desktop system now 26% higher in terms of average FPS. Even its 1% low is now higher than the laptop’s average FPS, though this is one of the biggest differences seen in all 17 titles tested.

The margins don’t change too much if we enable ray tracing, though the performance dips substantially on both. If we instead use DLSS without ray tracing, well the differences are much less pronounced.

Remember at 1440p stock the desktop system was 26% higher, now with DLSS it’s only half a frame faster, basically no difference at all. F1 2021 was doing better on the laptop in terms of average FPS at both resolutions, however it’s only a slight lead, and the 1% lows from the desktop were consistently better, so again overall a more stable experience there which could be down to the higher powered processor.

Watch Dogs Legion on the other hand had the biggest difference out of all 17 games tested in favor of the desktop graphics card. The 1% lows were higher than even the average FPS that the laptop was able to produce at either resolution. At 1080p the desktop was reaching 33% higher average FPS, then a massive 52% gain at 1440p.

I think this might be due to the VRAM differences, as the game was warning the laptop that 6 gigs may not be enough for ultra settings. Like most of the other games that have DLSS support, the desktop also sees a further higher boost with this enabled. At 1440p for instance, the laptop only gets a 4 FPS boost with DLSS, while comparatively the desktop was about 15 FPS higher now. Microsoft Flight Simulator was one of the better results in favor of the desktop graphics card, reaching a 12% higher average frame rate at the higher 1440p resolution and 10% higher at the lower 1080p resolution, though again, there’s a larger difference in the 1% lows. Metro Exodus should be fairly GPU heavy at ultra settings, but regardless of resolution this game had one of the smallest differences out of all 17 tested.

The desktop graphics card was ahead, but with just a 1 FPS difference or so in the averages and a 2-3 FPS improvement in the 1% lows, it’s not a significant amount at all. Assassin’s Creed Valhalla saw above average gains with the desktop graphics, 11% faster at both resolutions compared to the laptop, granted in this game, I’d be surprised if the 5900X wasn’t helping things along. The differences were smaller in Fortnite, the desktop system was 8% ahead in average FPS at 1080p, but then at 1440p this lowers to a 6% lead, perhaps the extra CUDA cores in the laptop are able to get utilized better at higher resolutions in some games as we’re more GPU bound there. This game also has support for DLSS, and it’s a good way to further boost the performance on both machines, even 1440p at max settings on the laptop is running very nicely.

Instead of talking through the rest of the 7 games in detail, I’ll just quickly skip through them so we can get onto looking at the average differences of all games tested as that’s more interesting, just pause the video if you want a detailed look at any of these results.

Some games had basically no differences, while others had much larger changes, it depends on the game which is why I decided to test 17 titles to get a better rounded average. On average over all 17 games tested at 1080p, the desktop RTX 3060 was just 8% faster than the laptop, far less of a difference than I was expecting going into the testing. Although only F1 2021 was ahead on the laptop GPU here, 6 other games were less than 5% faster on the desktop, so for the most part I’d say the laptop is holding up quite well. That said some of the games like Watch Dogs Legion were able to see much more of an improvement with the desktop card, and I suspect in that specific case it’s due to the extra VRAM, as the game was warning about 6 gigs on the laptop not being enough for ultra settings.

Stepping up to 1440p and on average the desktop card is now 9% ahead of the laptop, but I think this is being held up by the massive gain in Watch Dogs Legion, which was now 50% ahead on the desktop card.

Higher resolutions generally benefit more from extra VRAM, and given we were already limited at 1080p in this game that makes sense. If we disregard Watch Dogs as an outlier, well the desktop card is just 6% faster on average. 4 games were technically doing better on the laptop now, though it’s close, and another 5 games were less than 5% faster on the desktop, a minor difference. Although the desktop card has its extra VRAM, the higher CUDA core count of the laptop GPU seems to help offset this.

The desktop system also seems to be using much more power when measured from the wall when compared to the laptop, more than double the total wattage – at least in this one specific game tested.

I think it goes to show that although desktop hardware performs better, it’s mostly due to it being able to use more power as there’s more space for cooling. Best case in our 1440p test, control was reaching 25% higher average FPS on the desktop compared to the laptop, but with more than double the power draw to achieve this there’s definitely an argument to be had for the laptop being more power efficient. This might come as a shock to some of you, but there’s more to life than just gaming, so before we get into the price differences let’s see how well both compare in some creator workloads. Blender was tested with CUDA, and the results were quite close together, which I suppose is down to the laptop 3060 having more CUDA cores than the desktop card, I’m guessing the higher power limit of the desktop card was enough to give it the win though. Likewise V-Ray was also tested with CUDA, and the desktop card was scoring 10% higher than the laptop graphics here.

Adobe Premiere was tested with the Puget Systems benchmark, but this test relies on the processor, so the 12 core 5900X in the desktop is definitely responsible for the 48% higher score here, again it’s unfortunate I didn’t have the 8 core 5800X instead. DaVinci resolve is typically more GPU heavy, but the processor definitely still factors in. The desktop card was scoring 40% higher than the laptop in this one, but again, the difference isn’t entirely due to the GPU difference. So the desktop 3060 is generally better as expected, but what about the price difference? Prices will of course change over time, so refer to the updates down in the description below.

Right now the Lenovo Legion 5 with RTX 3060 graphics can be picked up for $1560 USD, but this one has the 6 core Ryzen 5 processor and a 2 plus month shipping time – yikes. The 8 core 5800H model I’ve isn’t listed here at the moment, presumably also out of stock, though I have seen it on sale for about $1400.

Not particularly cheap, but at least you get a full package ready to go with CPU, GPU, screen, keyboard and RAM. For a desktop PC, well, it’s no secret that graphics card pricing has been all over the place. Nvidia’s MSRP for the RTX 3060 is $330 USD, but sorting from cheapest prices on Newegg we’re currently looking at $800 just for the GPU, or about half the cost of the laptop.

The rest will greatly depend on the other parts that you select, the 5900X I’ve used for example is another $570 alone, but honestly something like a 5600X for less money is considerably cheaper and probably a better match for the 3060. Of course if you already have a PC, just upgrading the GPU is an option that’s not available with a laptop I’m not going to go ahead and do an entire PC build because as mentioned, again there are plenty of different components you can put in here, which will affect the price. But once you factor in everything else like case, power supply, RAM, motherboard, storage for instance, you’re probably going to end up paying somewhat close to the laptop anyway.

So purely from a cost perspective I think both are probably somewhat comparable. Again I’m sure you can find cheaper desktop components and a cheaper 3060 laptop, I’m just speaking in general.

At the end of the day, to me personally, the cost difference isn’t even the major factor, it all comes down to portability and whether or not you actually need to take your machine with you. If you do need to take your machine with you to say work or school then good luck taking this thing with you, or otherwise maybe you just don’t have enough desk space to keep a full blown desktop PC sitting around.

The point is there are obviously scenarios where laptops have a clear advantage. Generally though if you do have the space for a desktop PC and you don’t need the portability then that’s what I would suggest. Mostly because it’s way easier to upgrade this in the future compared to a laptop.

If you want to upgrade a laptop CPU and GPU for instance, you basically have to buy a whole new machine because they’re soldered to the motherboard. Anyway I guess the point I’m trying to make is there are different options for different people with different use cases. Personally what I was most interested to see was that the laptop 3060 could hold up in many games compared to the desktop 3060. If you are someone that does need to buy a laptop I think you can rest somewhat easy knowing that the 3060 in the laptop isn’t all that much worse compared to the desktop variant.

As we’ve seen, the laptop is still able to offer a pretty great gaming experience, and I have also heard of some desktop PC buyers resorting to buying laptops because it’s all that they’ve been able to find in stock.

So which would you pick, laptop or desktop? Let me know down in the comments below, and check out these videos next if you want to get an idea of how well last gen laptop GPUs compare against desktop GPUs.

I think it’s still very interesting because that last gen hardware still performs great and is more commonly available compared to these newer 3060s. Get subscribed if you’re new to the channel for more comparisons like this one, and come and join me in Discord and get behind the scenes videos by supporting the channel on Patreon..

Read More: RX 6600 vs RTX 3060 – 18 Game Test 1080p & 1440p

Leave a Reply

Your email address will not be published. Required fields are marked *