AMD vs NVIDIA | Which GPU is Better in 2023?

In this article, we have described the features on which basis you can understand that which GPU wins in the fight between AMD vs NVIDIA. So, if you’re putting together a general or gaming PC, you’ll have to choose between the two GPU giants. Both firms provide GPUs that power the top graphics cards, and they compete for first place in our GPU benchmarks hierarchy.

Of course, AMD versus Nvidia isn’t the only choice you’ll have to make when putting up a PC. You’ll also have to decide between AMD and Intel CPUs. However, our focus here will be on graphics, and we’ll look at performance, features, drivers and software, power and economy, cost, and other factors.

The AMD vs. Nvidia flame wars have been raging since the late 1990s, with Nvidia now winning in several measures in the GPU area. Its graphics cards, for example, account for the majority of GPUs on the Steam Hardware Survey, and in terms of pure financial value, Nvidia is about three times as valuable as AMD (with a big portion of AMD’s resources devoted to CPUs).

Be sure to check out our many other buying guides, including the Best GPU for i7 10700K, Best Graphics Cards for 1080p 144hz, Best Graphics Cards for Ryzen 9 5950X, Best GPU for Ryzen 7 3700X, Best Graphics Cards for Ryzen 9 5900x, Best Graphics Cards for Ryzen 7 5800x, and Best Graphics Card under $300

But we’re not concerned with the distant past or money. We seek to uncover a winner in the current GPU war between AMD and Nvidia. That essentially means looking at AMD Big Navi and Nvidia Ampere graphics cards and perhaps one of these days, Intel Arc will be added to the mix.

Throughout this study, it is critical to keep the overall picture in mind. We’re not only looking for the quickest GPU, the most power-efficient GPU, or the best GPU for the money. We’ll look at everything in each category, from cheap to mid-range to high-end and extreme GPUs, as well as the technology behind the GPUs. We’ll proclaim a winner today, but it’s not the end of the conflict. It’s more like to having the heavyweight GPU title: a win today does not guarantee that your opponent will not return leaner and meaner next year.

With everything out of the way, let’s put on our boxing gloves and go round by round with AMD against Nvidia.

AMD vs NVIDIA: Gaming Performance

Faster GPUs have allowed game developers to construct more rich and intricate environments for decades. While both AMD and Nvidia offer everything from affordable GPUs to high-end solutions, Nvidia has a tiny overall advantage owing to the beefy GeForce RTX 3090 Ti.

Beyond the pole position, it’s a tighter race. In our GPU benchmarks hierarchy, AMD’s RX 6950 XT takes the top place at 1080p and 1440p, while Nvidia only wins at 4K or in ray tracing games. Four of the top ten fastest GPUs are AMD and six are Nvidia, but this is due to quantity rather than fundamentally distinct GPUs – the RTX 3080 12GB combines a little of the RTX 3080 Ti and a bit of the RTX 3080 and blends them together to create a card that matches the 3080 Ti at a lower price point.

Of course, the primary benchmarks only look at games that use APIs and settings that work on all GPUs, thus ray tracing and DLSS aren’t included in the findings. We also didn’t include any FSR findings, and CPU constraints are undoubtedly a factor at lower resolutions. Here are the revised 2022 performance rankings, which show the overall performance from eight games and four settings/resolution combinations in our normal test suite, as well as four additional charts for our ray tracing suite, which employs six difficult DXR (DirectX Raytracing) titles.

Also, check out our separate post on Install or Update NVIDIA Drivers

Nvidia’s RTX 3090 Ti is too expensive, while the GeForce RTX 3080 and Radeon RX 6800 XT offer a more realistic depiction of performance. In standard rasterization games, AMD takes a tiny lead, whereas Nvidia easily surges ahead frequently by a significant margin once ray tracing and/or DLSS are enabled. Nvidia has a little advantage at the top of the performance ladder, but that isn’t the only factor to consider.

When it comes to the mainstream market ($400, give or take), things become a little complicated. AMD’s Radeon RX 6700 XT competes with the GeForce RTX 3060 Ti, with slightly greater performance if we discount RT and DLSS. In our evaluation, we referred to it as a net tie. While AMD’s card technically costs 20% more than Nvidia’s offering, pricing currently favours AMD ($480 vs. $500 at the time of writing).

Older GPUs are typically outperformed by newer models; thus, the optimum combination would be the RTX 3050 vs the RX 6600, both of which can be bought for roughly $300. (Never mind MSRPs). In regular gaming performance, AMD’s GPU outperforms the 3050 by 25-35%, and while ray tracing favors Nvidia GPUs by 7-15%, it’s significantly less important in this area.

What about low-cost options, such as items under $200? Unfortunately, we must question where all of the decent budget GPUs have gone. The RX 6500 XT and RX 6400 are less than $200, however they perform lower than the previous RX 5500 XT 8GB and GTX 1660, instead competing with the GTX 1650 and 1650 Super. AMD’s cards are very simple to buy brand new, however Nvidia’s GTX 16-series chips are now three years old, and supply and availability can be inconsistent. Nonetheless, Newegg offers the GTX 1650 Super for $200, virtually tying the budget segment.

Winner:

Because there are so many possible outcomes, this is simply too close to call. Nvidia takes the lead in 4K and ray tracing performance, while AMD takes the lead in conventional 1080p and 1440p gaming, as well as performance at comparable high-end and mid-range price points. However, DLSS works in Nvidia’s advantage, and while FSR 2.0 is pretty competitive, it works on any GPU.

AMD vs NVIDIA: Power Consumption & Efficiency

Prior to AMD’s Navi, Nvidia’s GPU power efficiency was clearly superior. But Navi changed everything, and Big Navi has boosted AMD’s efficiency even further. Navi began to bridge the gap by using chips made with TSMC’s 7nm Fanfest technology and a new design that offered 50% greater performance per watt. Except that it was so far behind that even a 50% improvement wasn’t enough to make up for the inefficiency.

However, Nvidia’s Ampere design pushes greater clocks at the expense of efficiency, whereas AMD’s Big Navi benefits from the Infinity Cache. As a consequence, Ampere and Big Navi are almost evenly matched.

We evaluated all of the current and recent graphics cards from both brands using Powenetics technology to record the true graphics card power utilization of GPUs. We also examined third-party cards from both sides, but we’ll keep the charts as close to the reference designs as feasible.

While power consumption favors Nvidia’s previous GPUs by a large amount, the disparities on the next generation hardware are the opposite. The RX 6950 XT consumes the most power, followed by the RTX 3080 12GB (custom card), and finally the RTX 3090, 3080 It, and 3080. Before we go to prior generation GPUs, we’ll look at AMD’s RX 6900 XT and 6800 XT.

Outside of the extreme performance range, it is more difficult to designate a winner. AMD’s RX 6700 XT consumes somewhat more power than the RTX 3060 Ti and slightly less than the RTX 3070, but they all fall within the 10W range. When comparing the RTX 3060 and RTX 3050 to the RX 6600 XT and RX 6600, power consumption roughly tracks performance.

Go through our epic guide on install a new graphics card

The RX 6500 XT consumes more power than the GTX 1650 among cheap GPUs, but the RX 6400 is the only fully contemporary GPU that does not require at least a PCIe 6-pin power adapter.

Winner:

AMD Power and efficiency are quite similar when focusing just on the current generation AMD Big Navi and Nvidia Ampere GPUs. AMD takes a little advantage at the top, the middle is tied, and AMD also wins in the budget sector, giving it the overall win. AMD benefits from TSMC’s N7 process, whereas Nvidia’s use of Samsung 8N likely accounts for lower overall efficiency.

AMD vs NVIDIA: Technology Highlights

The majority of the functions offered by AMD and Nvidia appear to be identical, albeit implementations differ. Both now support ray tracing, which allows for some excellent effects, but it isn’t necessary for a good gaming experience. FidelityFX Super Resolution (FSR) doesn’t look as well in the FSR 1.0 implementation, and FSR 2.0 isn’t widely accepted yet in either case, FSR works on AMD and Nvidia GPUs, as well as Intel integrated graphics. That’s really a plus for AMD because it doesn’t lock users into its hardware, but in practice, there are a number of games with DLSS support, and that can be a pleasant addition.

There are further considerations. G-Sync competes with FreeSync, Radeon Anti-Lag competes with Nvidia’s ultra-low latency mode with Reflex, Radeon Super Resolution (RSR) operates similarly to Nvidia Image Scaling (NIS), and there are several additional areas where features effectively match up

Supporting the same APIs and similar hardware characteristics, however, does not make things identical. Ampere and RDNA2 also support mesh shaders and variable rate shading (VRS), as well as a few other DirectX 12 Ultimate capabilities. However, even without DLSS, Nvidia’s ray tracing performance is far greater than AMD’s.

While FreeSync and G-Sync may appear to be interchangeable on the surface, the best G-Sync displays nearly always have superior quality and lower latency than FreeSync displays. The same may be said about AMD’s anti-lag technology and Nvidia’s ultra-low latency and Reflex: In principle, they are similar, but in reality, Reflex implementations win.

Another significant factor is video encoding and decoding, and Nvidia clearly wins here. Higher-quality encodes and lower GPU consumption are also advantages of the Turing and Ampere codecs. Nvidia’s newer GPUs eliminate the requirement for CPU-based video encoding. AMD’s low-cost Navi 24 processors are also lacking in crucial video encoding functionality.

There is one area where AMD clearly has an edge, however it may potentially be to AMD’s harm. TSMC’s N7 process, which AMD utilizes for RDNA2 (and RDNA and Zen 3 and the PS5/XSX) processors, obviously outperforms Samsung’s unique 8N (actually just an enhanced version of Samsung’s 10LPP process). The hitch is that many other businesses want to benefit from TSMC’s goodness as well – AMD, Apple, Nvidia GA100, Qualcomm, and even Intel, among others, employ TSMC. This creates an issue when TSMC’s capacity is unable to fulfil the requests of all of those enterprises.

Winner:

While AMD and Nvidia appear to share most functionalities, Nvidia’s implementations are typically superior – and more expensive. G-Sync, Reflex, DLSS, and NVENC all perform somewhat better than AMD’s counterparts.

NVIDIA vs AMD: Drivers and Software

It is tough to pick a clear champion in the drivers and software area. Several customers previously reported black screen issues with AMD drivers on RX 5000 Navi series GPUs, while others reported no problems. As far as we can determine, newer drivers have resolved these issues. Nvidia drivers aren’t perfect either, and bugs arise for both companies depending on the game and hardware. Is one firm, however, performing better with drivers than another?

AMD makes a big deal about their annual driver update. The Radeon Adrenalin 2020 drivers combined everything under one broad umbrella in an attempt to simplify things, however it might be perplexing at first if you’re used to the earlier drivers. AMD frequently skips WHQL (Windows Hardware Quality Labs) testing, which means fewer hoops to pass through and maybe more flaws slip through, but Microsoft’s assurance of a minimal level of operation doesn’t mean much for gaming reasons.

You can usually expect at least one new AMD driver every month, and typically more if there are large game releases. The timeline for Nvidia’s drivers is similar. For significant game releases or new graphics card hardware, you’ll receive fresh drivers. Except for occasional hotfixes, Nvidia’s releases are entirely WHQL certified, and the company also offers a distinct Studio Driver branch for content producers.

One significant distinction between AMD and Nvidia drivers is that Nvidia has two distinct user interfaces. The Nvidia Control Panel covers resolutions and specific graphics settings, whereas GeForce Experience manages game optimizations, driver updates, and additional features like as ShadowPlay, Ansel, NIS, and others. To utilize GeForce Experience, you must first log in and solve a captcha prompt, which I’ve done more times than I’d want to count. Simply say no to data mining and drivers.

Follow our guide to What Graphics Card Do I Have

Winner:

We like AMD’s unified driver approach since it reduces the number of interfaces to visit, but there’s simply so much in the current versions. Nvidia’s Q&A is undoubtedly superior, however both sides have lots of flaws and difficulties. However, quantifying drivers is a highly subjective endeavor, so we’re calling this one a tie as well.

AMD vs NVIDIA: Costs and Availability

In the AMD vs. Nvidia debate, who provides the better value? The last two years have been a bit of a joke (not a humorous one), but things are now much better. All of the current graphics cards are now available at retail, with several models selling at or below MSRP. However, significantly more of the “many” selling at or below MSRP use AMD GPUs than Nvidia GPUs, with only the excessively expensive RTX 3080 Ti and beyond selling at or below MSRP from Team Green.

Our GPU pricing index considers eBay pricing as an option; however, it is no longer necessary. In general, AMD presently offers the best value, with the RX 6600, RX 6600 XT, RX 6650 XT, RX 6700 XT, and RX 6750 XT all placing towards the top of the FPS per dollar lists. The RTX 3060, 3050, and 3060 Ti from Nvidia aren’t bad options, but you’re paying approximately 15-25% more in terms of FPS per dollar (and getting DLSS for that money).

The RTX 3090 Ti is the undisputed poorest bargain at the top of the pricing scale, costing more than $1,800 and giving only marginally higher performance than the RTX 3080 Ti. AMD’s RX 6950 XT is less expensive and competes effectively against the 3080 Ti, but as we move down the pricing range, AMD’s advantage grows significantly.

The RX 6600 now holds the top spot, with roughly 0.23 frames per dollar spent. Nvidia’s best-ranking GPUs are from the previous generation Turing RTX 20-series, with the RTX 2060 scoring 0.196 FPS/$ and the highest RTX 30-series component scoring 0.176 FPS/$. At 0.176 FPS/$, AMD’s RX 6500 XT is squarely in the fight for a budget card.

Final Thoughts

Nvidia graphics cards are the best choice for ray tracing, streaming, video creation, and professional applications. Its cards outperform AMDs in terms of ray tracing capabilities, and more Nvidia cards are compatible with 3D modelling applications than AMD cards. If you don’t need ray tracing and just want to push frame rates, AMD will provide higher performance at the same price.

The gap between Nvidia and AMD is substantially narrower than it used to be. AMD’s user interface is superior, and it is continually introducing new features and upgrades to increase performance and match Nvidia’s feature set. Monitor synchronization, upscaling, ray tracing, internet streaming, and noise reduction are all supported by both manufacturers.

The implementation of these functions by Nvidia is smoother, but most of us don’t need them to begin with. If you want to save money, AMD cards make more sense, and you should consider the tier of card you require rather than the brand.

The Nvidia RTX 3060 and AMD RX 6600 XT are both fantastic picks for 1080p gaming. It’s recommended upgrading to the Nvidia RTX 3070 or AMD RX 6700 XT for 1440p gaming. If you want to push 4K frame rates or use ray tracing, the RTX 3080 is a good investment. The AMD RX 6900 XT can handle 4K gaming, but it struggles with ray tracing and won’t save you much money over comparable Nvidia cards.

Check out our separate post on Why Are Graphics Cards So Expensive

FAQs for AMD vs NVIDIA