AMD vs Nvidia. If you're building a gaming PC, you'll inevitably be faced with choosing between the two GPU heavyweights. Both companies make GPUs that power the best graphics cards, though Nvidia currently occupies the top positions of our GPU benchmarks hierarchy. AMD vs Nvidia isn't the only decision you’ll need to make when building a PC, of course. You'll also need to choose between AMD vs Intel CPUs. Our focus here will be on graphics, however, and we'll be looking at performance, features, drivers and software, power and efficiency, pricing and more.
The AMD vs Nvidia flame wars have been cooking since the late 90s, back when AMD's current GPU division consisted of the ATI brand. AMD is a much older company, with its roots stretching back to the late 50s (May 1, 1959 is the actual founding date). Nvidia is about half as old by comparison, but the plucky young upstart has become the king of the graphics industry. In terms of pure finances, Nvidia is worth roughly twice as much as AMD, and a large chunk of AMD's resources are devoted to CPUs as well.
But we're not interested in the distant past or finances. We want to find a winner in the current battle of AMD vs Nvidia GPUs, and we're also looking forward to seeing how things might change when AMD Big Navi takes on Nvidia Ampere later this year (with Intel Xe Graphics perhaps hoping to spoil the party).
It's important to keep the big picture in view throughout this analysis. We're not just focusing on the fastest GPU, or the most power-efficient GPU, or the best bang-for-the buck GPU. We'll consider all of the factors in each category, from budget to mid-range to high-end and extreme GPUs, along with the tech behind the GPUs. We will declare a winner today, but of course this isn't the end of the war. It's more like owning the heavyweight GPU title: A victory today doesn't mean your opponent won't come back leaner and meaner next year.
With that preamble out of the way, let's pull out the boxing gloves and go the rounds with AMD vs Nvidia.
AMD vs Nvidia: Gaming Performance
For decades, faster GPUs have enabled game developers to create increasingly detailed and complex worlds. While you can find everything from budget GPUs to high-end offerings from both AMD and Nvidia, when it comes to outright performance, Nvidia has a clear overall lead. If you look at our GPU benchmarks hierarchy, you'll see that (not counting Titan cards, which aren’t really aimed at just gamers) Nvidia holds the top five spots. The best AMD can do is sixth place, with the Radeon VII and RX 5700 XT nearly tied for overall performance.
That makes this a pretty easy win for Nvidia at the top of the performance ladder, but that's not the only category to consider. Once we reach the $350 mark, AMD GPUs become far more competitive. We've looked at AMD's RX 5700 XT vs Nvidia's RTX 2060 Super elsewhere, and gave the RX 5700 XT a slight edge overall. It's generally faster and costs less, though it does use more power and lacks support for ray tracing. Stepping down another notch, the RX 5600 XT vs RTX 2060 is pretty much a straight-up tie. Nvidia still wins with features, but performance and other metrics are extremely close.
What about in the budget category? RX 5500 XT vs GTX 1660 goes back to Nvidia. We've since retested both performance and power of both GPUs, however, and additional testing only proved that Nvidia now wins the power and efficiency category. We'll dig into the details more below, but it's a 4-to-1 lead, even if many of the categories are close. Nvidia is a touch faster and uses a bit less power, for roughly the same price.
What about cards that cost even less, like the Radeon RX 5500 XT 4GB, Radeon RX 570 4GB, and GTX 1650? What does AMD vs Nvidia performance look like at the bottom of the pricing spectrum? RX 570 4GB is still the cheapest 'decent' graphics card, with prices starting at $120. It's a bit faster than a GTX 1650 in most games, but slower than a GTX 1650 GDDR6, while using roughly twice the power. It's also 20% slower than the newer RX 5500 XT 4GB.
Winner: Nvidia Across a large suite of games, Nvidia wins in most categories. AMD can hold its own with the RX 5700 and 5700 XT, but it can't beat the 2070 Super or above and basically matches the old GTX 1080 Ti. Even in the lower-priced categories, a traditional stronghold for AMD, Nvidia's current lineup generally delivers superior performance at equivalent prices. And that's not even factoring in ray tracing support on the RTX cards, something which AMD won't offer until Big Navi later this year.
AMD vs Nvidia: Power Consumption and Efficiency
Prior to AMD's Navi, at least for the past six or more years, the competition between AMD and Nvidia in terms of GPU power efficiency was decidedly in favor of Nvidia. But Navi changed all that, right? Using chips built with TSMC's 7nm FinFET process and a new architecture that delivered 50% better performance per watt, it could close the gap. Except, it was so far behind that even a 50% improvement didn't fully address the efficiency deficiency.
Using Powenetics hardware to capture the real graphics card power use of GPUs, we recently retested all of the current and recent graphics cards from both companies. Navi is certainly better than any of AMD's pre-Navi chips, but Nvidia still wins overall, even using GPUs that are 18 months old built on TSMC's previous-generation 12nm node. Perhaps Big Navi will change that, but Big Navi needs to take on Nvidia Ampere, not Turing.
Let’s look at the different GPU target markets. In the extreme performance realm, Nvidia's RTX 2080 Ti and RTX 2080 Super use a lot of power but don't really have any direct AMD competition. AMD's Radeon VII and older Vega 64 basically use just as much power, if not more, but deliver substantially less performance. AMD's RX 5700 XT splits the difference between the RTX 2070 and RTX 2070 Super, while using slightly more power than the 2070 Super.
AMD's mainstream offerings claim victories at least, with the RX 5700 beating the RTX 2060 in performance using the same amount of power, and coming in just behind the 2060 Super while using less power. It's not a major win, but it's better than a loss. The 5600 XT likewise delivers slightly better than RTX 2060 performance while using slightly less power (and yes, that's with the 14 Gbps GDDR6 update applied).
Budget Navi 14 cards don't do quite so well. The Radeon RX 5500 XT 4GB and 8GB models consume nearly the same amount of power, about 125W — that's while gaming, not in a worst-case test like FurMark where they hit 170W. By comparison, Nvidia's GTX 1660 Ti and GTX 1660 Super use a bit less power but perform up to 20% faster. Alternatively, the GTX 1660 is a touch faster than the 5500 XT 8GB and uses 10W less power. It's not a major difference, but it's still a win. Similarly, dropping down to Nvidia's GTX 1650 line, the GTX 1650 Super edges out the RX 5500 XT 4GB by a hair in performance while using 25W less power, while the GTX 1650 GDDR6 shaves off another 20W but also drops 17% in performance.
Winner: Nvidia Focusing just on the current generation AMD Navi and Nvidia Turing GPUs, power and efficiency are relatively close. Nvidia wins at the top and bottom of the price and performance spectrum, while AMD edges out Nvidia in the mainstream sector. The real concern here is that Nvidia still gets the win even while using GPUs built using previous-generation lithography. Ampere could be a monster, though there's nothing to stop AMD from further improving efficiency with RDNA 2.
AMD vs Nvidia: Featured Technology
Let's just cut straight to the point: Do you actually need ray tracing to get good graphics in a PC game? Obviously the answer is no, but ray tracing does allow for some nice effects. This is the big difference between AMD and Nvidia features right now. Everything else is secondary. G-Sync takes on FreeSync, Radeon Anti-Lag goes up against Nvidia's ultra-low latency mode, and plenty of other areas are basically tied. But Nvidia offers GPUs with ray tracing hardware and AMD does not, at least not yet (it's coming with Xbox Series X and PlayStation 5, along with Big Navi).
Okay, ray tracing isn't the only difference. Nvidia has also supported Variable Rate Shading (VRS) since the launch of Turing GPUs, and DLSS (Deep Learning Super Sampling) uses the Tensor cores found in Nvidia's RTX GPUs. Turing also supports mesh shaders and some other features that are all part of the DirectX 12 Ultimate spec, and Turing RTX GPUs are fully compliant with the Vulkan Ray Tracing spec. If we're talking about which company has pushed out more new graphics features over the years, features that actually become part of the greater graphics ecosystem, it's Nvidia. That’s despite major pushes that never really caught on in a big way, like PhysX and 3D Vision.
AMD isn't sitting idle, and some of Nvidia's tit-for-tat features were in response to AMD features, but the same is also true the other way around. Would we have FreeSync if Nvidia hadn't first developed G-Sync? AMD also offers a lot of new features via open source rather than using proprietary closed designs. PureHair was open source, HairWorks was not. FidelityFX / Contrast Aware Sharpening (CAS) is open source and works with all GPUs. Ansel and various Nvidia features require an Nvidia GPU. AMD's Mantle API from way back when also eventually helped to spur DirectX 12 and Vulkan features. But that was five years ago, and this is now.
A separate aspect of the technology and features in the current GPUs is the manufacturing process. AMD bet big on TSMC's 7nm FinFET node last year, and it certainly paid off. It helped the company's Zen 2 architecture-based Ryzen CPUs surpass Intel's CPUs in many respects, and it definitely contributed to gains in efficiency on the Navi GPU front. AMD certainly earns some points here for being more aggressive in adopting newer manufacturing nodes — and it could have bitten AMD in the butt, which has happened in the past. Both AMD and Nvidia are expected to have new chips using 7nm by the end of the year, though, and the manufacturing process doesn't necessarily decide how good a GPU is (outside of potential gains in efficiency, performance, and/or features).
Winner: Nvidia Once the next-generation hardware arrives later this year, we might be at feature parity again, but right now Nvidia is leading the charge on ray tracing and other recent enhancements like DirectX 12 Ultimate and Vulkan RT. Ray tracing is clearly a long-term play and likely could become the standard approach to graphics rendering in games over the next decade or so. Features like Tensor cores for real-time denoising and DLSS open the door for ray tracing on lower-spec GPUs.
AMD vs Nvidia: Drivers and Software
Trying to determine a clear winner in the drivers and software category is difficult. Quite a few people have encountered black screen issues with AMD drivers on RX 5000 Navi series GPUs, while others haven't had any difficulties. Newer drivers have supposedly fixed these problems, but some user complaints continue. Nvidia drivers aren't foolproof either, and depending on the game and hardware, issues crop up for both companies. But is one company doing better with drivers?
AMD makes a lot of noise about its yearly driver overhaul. The Radeon Adrenalin 2020 drivers consolidated everything under one large umbrella, aiming to simplify things, though it can be confusing at first if you're used to the older drivers. AMD has pushed out twelve main driver updates this year, of which three were WHQL certified. WHQL, or Windows Hardware Quality Labs, is Microsoft's team that verifies a minimum level of functionality. You can generally count on at least one new AMD driver per month, often more if there are major game launches.
Nvidia's driver schedule follows a similar cadence; you'll get new drivers for major game launches or new graphics card hardware. So far, there have been nine major driver releases in 2020 that are all WHQL certified, plus five (maybe more) hotfix driver updates. Nvidia also has a separate Studio Driver branch for content creators, which has had four releases this year.
One of the big differences between AMD and Nvidia drivers is that Nvidia has two separate user interfaces. The Nvidia Control Panel handles things like resolutions and certain graphics settings, while GeForce Experience tackles game optimizations, driver updates, and extra features including ShadowPlay, Ansel, and more. Annoyingly, you have to log in and solve a captcha prompt to use GeForce Experience, which is something I've done more times than I’d ever want to count. Just say no to data mining plus drivers.
Winner: Tie We prefer AMD's unified driver approach, as it's one less interface to navigate, but there's just so much stuff in the 2020 release. Moreover, while I haven't personally experienced any repeatable issues with AMD's drivers, Nvidia's Q&A is arguably better. The black screen complaints with Navi went unfixed for over half a year, and there may still be lingering problems. AMD's drivers have improved substantially since the days of the Catalyst Control Center, but sometimes less is more. Quantifying drivers ends up being an incredibly subjective affair, however, so we're calling this one a draw.
AMD vs Nvidia: Value Proposition
Who offers the better value in the battle of AMD vs Nvidia? So far, Nvidia has chalked up wins in three categories and a tie in one, but price is the great equalizer. AMD may not have the fastest or most efficient GPUs, but it will often sell you competitive performance at a lower price. Or at least, that used to be the case; things may have changed. Here's how pricing breaks down looking at the current GPUs.
At the extreme end of the pricing scale, AMD doesn't even bother to compete. Forget the $2,500 Titan RTX; even the $1,200 RTX 2080 Ti carries an obscene price tag! For that matter, the RTX 2080 Super and RTX 2070 Super also cost substantially more than anything AMD currently sells that's worth buying (no, the Radeon VII doesn't count, sorry). If you want the fastest GPU around, sure, Nvidia will sell that to you. Arm and leg donations go in the box to the right. Technically, Nvidia is still the best value in extreme GPUs, if only by AMD's abdication of any claims to the throne.
The high-end market, starting at around $350 and stretching to $500, is where things get much more interesting. AMD's most-expensive GPU that you should even consider buying right now is the RX 5700 XT, nominally priced at $400. Prior to the impact of COVID-19, prices were dipping as low as $350, and if you're willing to deal with mail-in rebates you can still nab an RX 5700 XT like the ASRock 5700 XT for $360 ($380 without the rebate). You won't get ray tracing support, but the 5700 XT ends up being about 13% faster than the RTX 2060 Super on average, with 2060 Super prices starting at $400. Alternatively, the RTX 2070 Super is about 5% faster than the 5700 XT, except prices jump to $500. AMD clearly wins the value proposition for the high-end sector.
What about the $230 to $350 mid-range sector? There are five solid propositions here, perhaps more, with a pretty wide gap in performance and features. Looking at the amount of performance you'll get per dollar spent, the RX 5700 is about 22% faster than the RX 5600 XT (yes, the 14 Gbps memory variant), which in turn leads Nvidia's RTX 2060 by 5%. The 2060 is around 17% faster than the 1660 Ti, which is only 1-2% faster than the GTX 1660 Super. Ultimately, RX 5600 XT vs RTX 2060 gives Nvidia a win, mostly because of features as performance is basically tied, but the RX 5700 is actually our recommended pick at the top of the $300-ish range. If you only want to spend around $250, RX 5600 XT vs GTX 1660 Ti isn't even close (AMD wins), and while the 1660 Super does better, the RX 5600 XT ultimately gets the nod — even with a price that's about 15% higher. The mid-range value category nets AMD another win.
Budget GPUs priced below $230 have jumped in pricing since the previous generation (i.e., GTX 1050 vs RX 560), but performance is also better. Sure, there are still $120-$130 RX 570 4GB cards available, but the new RX 5500 XT 4GB is about 25% faster, uses half as much power, and costs just $30 (23%) more. You could call it a wash, but we recommend skipping the 570 4GB at this point. However, the RX 5500 XT 4GB isn't actually our pick among budget GPUs, since the GTX 1650 Super is about the same performance and, according to our revised power testing, it's also more efficient: GTX 1650 Super uses 25W less power than the RX 5500 XT 4GB. That's a small win for Nvidia. At the top of the budget spectrum, things get messy. GTX 1660 Super is 20% faster than the RX 5500 XT 8GB but costs 28% more. The vanilla GTX 1660 GDDR5 on the other hand is only 3% faster and costs about 10% more, so AMD's RX 5500 XT 8GB is a better value. Overall, there's a lot of flux in pricing on the budget GPU front, so we're going to call the budget segment a tie — Nvidia wins the ultra-budget with the GTX 1650 Super, and AMD has an edge if you're willing to spend a bit more money.
Winner: AMD After looking at the whole spectrum, AMD ends up being the value leader. There's a lot of subjectivity in here, but we give more weight to the mid-range and high-end price categories since that's where a lot of our favorite GPUs reside. As much as we love the performance of the RTX 2080 Super and RTX 2080 Ti, we know most gamers are far more likely to buy a $250-$350 graphics card.
|Round||Nvidia GeForce||AMD Radeon|
|Drivers and Software||✗||✗|
AMD vs Nvidia: Bottom Line
With an overall score of four to two, Nvidia continues to reign as the GPU champion of the world. Cue the fireworks. AMD lands some solid blows, particularly as a value proposition, and we're certainly interested in seeing how Big Navi and Ampere change things up later this year. But if you're looking at the big picture, including performance, efficiency, features, and the underlying technology, Nvidia is in the lead — and you could even argue that Nvidia deserves the drivers category as well, though we pushed on that one.
However, when you’re searching for the best graphics card for your needs, that could be an AMD-powered graphics card, particularly if you’re looking for the best bang for your buck. The answer for any given consumer really depends on their needs and current pricing.
The unfortunate side effect of Nvidia's pole position is the current prices we see on the highest-performance cards. With no real competitor for the extreme performance category for the past five years, we've watched Nvidia's steady climb up Mt. Expensive. Hopefully the RTX 20-series launch represented the summit and things trend downward again, but that's all contingent on AMD continuing to put the pressure on Nvidia. Another 50% improvement in performance per watt with RDNA 2 could close the gap, or Nvidia could go for a shock and awe campaign with its shift to 7nm lithography and try to blow AMD off the mountain.
For now, Nvidia continues to dominate. There are other aspects of GPUs we didn't delve into, like supercomputers, data centers, and deep learning hardware where Nvidia wins again. Then there’s the professional market, where Nvidia Quadro cards constitute a seemingly unassailable majority of users. AMD may have taken the lead in the CPU race over Intel, but it still has an uphill battle in the GPU war. Perhaps its profits on the processor side will give the company the muscle it needs to gain ground and, at some point, overtake Nvidia. Regardless of where your allegiance lies, competition is important to drive performance and features forward while bringing prices down to affordable levels.