Nvidia kept the launch of the GeForce RTX 5060 – successor to the popular GeForce RTX 3060 and RTX 4060 series – for Computex. It wasn’t exactly quiet, but the launch was unusual – drivers enabling reviews weren’t released until Monday evening, after sales began. Likely due to backlash over the RTX 5060 still offering only 8GB of memory, four years after the 12GB RTX 3060. As a result, it’s faced a fairly cold reception.
Final Thoughts
The main issue discussed with the 8GB RTX 5060 and RTX 5060 Ti is that 8 GB of memory might not be enough for newer games. If Nvidia had initially released both models exclusively with 16 GB of memory, priced just $50 higher than the 8GB versions, we could have avoided this debate. I’m not sure if Nvidia intentionally introduced the RTX 5060 solely in an 8GB version to push consumers, influenced by repeated claims that 8GB cards won’t handle new games, toward buying the 16GB RTX 5060 Ti for over 500 euros—implying that cheaper alternatives would only lead to disappointment.
In the previous generation, the price difference between the 8GB and 16GB versions of the RTX 4060 Ti was so large that the 16GB variant offered significantly worse value for money—unless your comparison focused heavily on games and settings where 8 GB of memory wasn’t enough. With the RTX 5060 and RTX 5060 Ti, the situation has changed. The gap between 8GB and 16GB RTX 4060 Ti models was $100 (€110), and even today, similar RTX 4060 Ti cards with 8 and 16 GB still differ in price by about €100. For the RTX 5060 Ti, however, the official price difference has shrunk to $50 (about €55), making the 8GB version not a good choice—unless you’re planning to play mostly esports titles at low settings.
If you want to save money and avoid paying too much for a card like the RTX 5060 Ti 16GB, it’s more sensible to go for the much cheaper RTX 5060 with 8 GB of memory. At its recommended price of $299, the saving compared to the 16GB RTX 5060 Ti is $130 (€140), which has a noticeable impact on the price/performance ratio.
Unlike others, I won’t try to convince you that you need to spend over 500 euros on a graphics 16GB card, or you won’t be able to play most new games well enough in a year or two. Claims that you won’t be able to enjoy new titles on an 8GB card in the near future are exaggerated and simply not accurate. That will only be true for certain games running at high detail settings and native resolution. BBut those same games probably won’t be playable at this level of detail on mid-range cards anyway—not just because of limited memory, but also because of insufficient GPU performance. The same applies to 4K resolution: even with a mid-range card, you can still play on a 4K monitor, but you’ll need to significantly reduce graphics settings and make use of FSR or DLSS.
I’m not talking about FSR 3 or DLSS 4 with frame generation. With less powerful cards, you can’t just set the game to run at 30 fps and expect frame generation to boost it to 50 fps—or even 80 fps with multi-frame generation. That approach doesn’t work because of input lag and frame latency. Even with frame generation enabled, you’ll still need to lower detail settings or apply DLSS and FSR more aggressively to push your base frame rate above 40 fps—so that, with frame generation, you reach smooth performance on a 60 to 75 Hz monitor.
With the arrival of DLSS and FSR, users gained another tool to get games running smoothly on less powerful cards while maintaining acceptable image quality—including titles that would otherwise be unplayable at native resolution. The possibilities for scaling performance are now greater than they were in the days of the 3GB and 6GB GTX 1060. Memory limitations are unlikely to become a widespread issue until consoles with more than 16 GB of shared system and graphics memory start to appear on the market—or in PC-only games focused on top-tier visuals.
When someone buys a graphics card for around 350 euros, it’s usually because they don’t want to—or can’t—spend an extra 100 euros. And if you’re going for a cheaper card, you can’t expect it to perform like a more expensive one. You’ll need to tweak graphics settings more often than you would with a higher-end model. If you don’t want to deal with that—or if you insist on running everything at max settings—then yes, go ahead and pay for the more expensive card with more memory.
Low on VRAM: What to Expect
You can usually tell when you’re running out of video memory by sudden drops to very low frame rates—either briefly or for extended periods. In some of the tested games at 4K settings, these drops even led to crashes, typically when ray tracing, upscaling, and frame generation were all enabled. The likely culprit is an apparent lack of memory.
For frame generation, I mostly used DLSS or FSR set to Balanced quality. To prevent crashes and avoid major frame rate drops even at 4K, it was usually enough to turn off frame generation and switch DLSS or FSR to a lower quality level. That won’t always be sufficient—some of the most memory-hungry titles weren’t included in the tests. However, even with those, lowering detail settings is often enough to make the game playable on an 8GB card without drastically reducing visual quality.
Buying a card like this for playing with ray tracing at 4K and maximum settings isn’t a good idea anyway. The GPU simply doesn’t have enough performance to ensure smooth gameplay in more demanding titles at high detail levels—even if the card has enough memory.
GeForce RTX 3060: Does 12 GB Still Matter When Performance Lags?
The RTX 3060 12GB was often mentioned as a card that wouldn’t run into memory capacity issues for years—and that holds true. In situations where games crash on the 8GB RTX 5060 due to memory limitations, the RTX 3060 handles them without crashing. And in titles that don’t crash, it also avoids the severe frame rate drops from 30 to as low as 5–15 fps. However, at the same settings where the RTX 5060 struggles with memory, the RTX 3060 still can’t deliver frame rates high enough for comfortable gameplay without lowering detail levels.
In scenarios where memory isn’t a limiting factor, the new RTX 5060 is often 40–66% faster than the RTX 3060—even at demanding native resolutions. Frame rates improve even further in games that support frame generationWhile the RTX 3060 doesn’t support DLSS frame generation, some games offer FSR as a basic workaround. However, even then, the base frame rate is often too low for frame generation to help—and the extra memory won’t make up for it. So yes, memory is sufficient, but GPU performance still requires lowering detail settings. RTX 5060 pulls ahead of the 3060 in many cases, particularly at lower resolutions, despite its smaller memory capacity.
Nevertheless, if you already own a 12GB RTX 3060, I wouldn’t recommend replacing it with the RTX 5060—unless you mainly play esports titles, where players typically prefer low-detail graphics. But if you plan to play more demanding new games with high-quality visuals, it makes more sense to spend a bit more on the 16GB RTX 5060 Ti, or at least hold off for a while to see if a refreshed RTX 5060 appears later this year. Nvidia is currently using 3GB memory chips only in laptops, likely due to limited availability, but if those chips make it into desktop cards, we could see an improved RTX 5060 with 12 GB of memory, which could put an end to the current memory dilemma.
Asus TUF GeForce RTX 5060 Gaming OC Ed.
The tested card represents the upper end of Asus’s RTX 5060 lineup, targeting users who value premium features. It is designed for brand enthusiasts willing to pay extra for superior design and top-tier equipment, despite not being the most cost-effective.
The card comes with heavily overengineered cooling—arguably more than necessary for a 155W GPU that isn’t hard to keep cool. On the upside, in quiet Q mode, you’ll mostly hear coil whine rather than fan noise. You’ll have a hard time matching this level of silence with system or CPU fans, which tend to be louder anyway.
In Q mode, the fans stay below 1000 rpm even under load, and they remain nearly inaudible—even in a passively cooled system at night. However, the large triple-slot cooler on a relatively efficient 155W card comes at a price—110 euros more than Asus’s baseline Dual RTX 5060, which, given the similar power draw, shouldn’t be much louder under gaming load—but we’ll only know for sure after testing.
Also, for similar money to the high-end TUF model, you can already find cheaper 8GB RTX 5060 Ti cards offering around 14% more performance. The TUF is worth considering mainly if low noise and temperatures are your top priorities and the price/performance ratio is less important. If you care about both noise and value, a Prime series card from Asus might be a more reasonable compromise. And there are certainly competing models that come close in noise levels while offering a lower price.
English translation and edit by Jozef Dudáš
Pros and Cons
+ Excellent thermal and acoustic performance
+ Dual BIOS, factory overclocking
+ Cooling is barely audible in Q mode
+ The cooler offers significant thermal headroom
+ Power limit headroom available for overclocking
+ Support for DLSS 4 and other proprietary Nvidia technologies
+ Improved ray tracing performance compared to the RTX 4060 and RTX 3060
– Much higher price than entry-level models
– Only 8 GB of memory
– Significantly worse price/performance ratio than entry-level RTX 5060s
– Entry-level 8GB RTX 5060 Ti models offer better performance for similar money
– Coil whine is quieter than on more powerfull cards, but still audible
Nvidia GeForce RTX 5060 graphics cards can be purchased in the e-shopu Datacomp.
⠀
