When it comes to building a PC for gaming, a GPU is one of the key items that you need to select carefully to achieve your desired performance and graphics. At the moment, NVIDIA and AMD are manufacturing GPUs with the best graphics cards. Here, we’ll be comparing the performance, features, power, efficiency, etc., of the GPUs by both companies.
Nvidia vs AMD
AMD is an old company while NVIDIA is just half as old as AMD. Still, NVIDIA occupies the top positions in GPU hierarchy due to its high-performing GPUs with high-resolution and frame rate. However, AMD has already announced to launch Big Navi GPU while NVIDIA would be launching Ampere GPU, thus, making their rivalry even bigger.
In this article, we’ll not just focus on the most efficient or the most powerful GPU by both companies, we’ll consider all the factors from power to graphics, efficiency, and performance. We’ll also discuss mid-range to high-range and extreme GPUs to show you a bigger picture of the GPUs manufactured by the two companies.
Gamers always love games with interesting details and complex scenarios. Such games can be played smoothly only with faster GPUs. While both NVIDIA and AMD are manufacturing both mid-range and high-end graphics units, NVIDIA always takes lead in providing outright performing GPUs.
When GPU performance is tested and compared for the two rivals, five NVIDIA GPUs outperform AMD GPUs. RX 5700 XT and Radeon VII nearly show similar performance in sixth place.
This makes NVIDIA as the top performer, however, that’s not the only category to consider. Once we reach the $350 benchmark, AMD GPUs become more competitive. In this regard, AMD’s RX 5700 XT shows a better overall performance when compared with NVIDIA’s RTX 2060 Super. It’s a fast-performing GPU with a much lesser price, though it consumes more power and doesn’t support ray tracing.
Similarly, AMD’s RX 5600 XT shows an extremely close performance when compared with NVIDIA’s RTX 2060, despite the latter possesses better features.
When it comes to comparing the budget GPUs of the two companies, NVIDIA’s GTX 1660 shows great power and efficiency along with performance when compared to AMD’s RX 5500 XT. This shows that for the same price, NVIDIA GPUs are a bit faster and consume less power.
When far more cheap GPUs are compared, RX 570 4 GB shows a bit faster performance compared to GTX 1650 in most of the games. However, it’s slower when compared with GTX 1650 GDDR6. Moreover, it consumes roughly twice the power compared to NVIDIA GPU.
So, in this category, NVIDIA wins due to its superior performance for both budget and high-end GPUs. Moreover, NVIDIA GPUs support ray tracing which is missing in AMD GPUs. Only the AMD’s Big Navi is expected to support ray tracing which is yet to be launched by the end of this year.
Power Efficiency and Consumption
When we compare power efficiency for the AMD and NVIDIA GPUs, NVIDIA is leading the competition for the past six or more years. But now with the launch of AMD’s Big Navi, things seem to change a lot. Because Navi consists of chips that are built with TSMC’s 7nm FinFET process. In addition to that, a new architecture would deliver 50% better performance per watt. Thus, the gap between the AMD GPUs and NVIDIA GPUs in terms of power efficiency could become quite close.
Otherwise, AMD was so far behind to NVIDIA that just a 50% improvement might not fully address the deficiency of power efficiency.
To capture the real power consumption of the graphics cards, we used the Powenetics hardware and retested all of the existing and latest GPU flagships from both companies. Although Navi is the best performer compared to all the existing AMD chips, NVIDIA still wins for providing power-efficient GPUs. Even the GPUs that are 18 months old and built on TSMC’s previous-generation 12nm node show better power efficiency compared to AMD GPUs.
Thus, Big Navi needs to take on NVIDIA Ampere in this regard. When the extreme performance realm of the GPUs is compared, NVIDIA’s RTX 2080 Super and RTX 2080 Ti consume a lot of power but don’t really have any direct competition with AMD GPUs.
When AMD GPUs are tested, Radeon VII and older Vega 64 use just as much power as NVIDIA’s GPUs but they deliver considerably lesser performance. AMD’s RX 5700 XT also splits the difference between NVIDIA’s RTX 2070 and RTX 2070 Super as it uses slightly more power than the latter GPU.
In this regard, AMD’s mainstream shows a better performance compared to its rivalry. The RX 5700 flagship shows a better performance than RTX 2060 while consuming the same amount of power. Moreover, it comes just behind the RTX 2060 Super in performance while consuming lesser power.
Likewise, AMD’s RX 5600 XT delivers slightly better performance than RTX 2060 while consuming slightly lesser power.
When Budget Navi 14 cards were tested for power efficiency, they didn’t do quite so well. Both the 4 GB and 8 GB Radeon RX 5500 XT models consume nearly the same amount of power while gaming i.e., about 125W. While in a worst-case test like FurMark, they consume 170W.
In comparison to the above-mentioned AMD GPUs, NVIDIA’s GTX 1660 Ti and GTX 1660 Super consume a bit lesser power but perform up to 20% faster. Moreover, the GTX 1660 is a bit faster than the RX 5500 XT 8 GB while it consumes 10W lesser power. Though it’s not a major difference, it still gives NVIDIA an edge over AMD.
When we go down to NVIDIA’s GTX 1650 line, the GTX 1650 Super slightly outperforms the RX 5500 XT 4 GB while consuming 25W lesser power. On the other hand, the GTX 1650 GDDR6 cuts power consumption by 20W but this also drops its performance by 17%.
When current generations of both companies are compared like NVIDIA Turing GPU and AMD Navi, power efficiency and consumption for both are relatively close. However, NVIDIA takes an edge over AMD for overall performance and price. While AMD’s mainstream outperforms NVIDIA in power efficiency.
NVIDIA GPUs which are built using previous-generation lithography also lead their AMD rivals in terms of power efficiency and performance. Ampere is also expected to be a monster in this regard, however, AMD can further improve its efficiency with the upcoming RDNA 2 flagship.
Featured technology is what every gamer prefers when buying a PC or laptop for playing games because it gives them a chance to play the latest versions of their favorite games. For example, ray-tracing in graphics cards allows some nice effects during the gameplay.
When it comes to AMD and NVIDIA, we see a big difference between the features of the GPUs manufactured by the two companies. Ray-tracing is one of the major differences between the NVIDIA GPUs and AMD GPUs which is not yet introduced in the latter. Moreover, G-Sync by NVIDIA takes on AMD FreeSync. While, AMD’s Radeon Anti-lag goes against NVIDIA’s ultra-low latency mode.
NVIDIA GPUs also support Variable Rate Shading (VRS) since the launch of Turing GPUs. While NVIDIA’s RTX GPUs come up with Deep Learning Super Sampling (DLSS) based Tensor cores.
Turing GPUs also support mesh shaders and some other features that are part of the DirectX 12 Ultimate specification. Furthermore, Turing RTX GPUs are fully compliant with the Vulkan Ray Tracing specification.
There are many areas where the performance of the GPUs from the two companies ties up. But if we talk about which company has introduced more new graphics features over the years that became part of the greater graphics ecosystem then it’s NVIDIA.
However, AMD isn’t just sitting idle. As some of NVIDIA’s features were in response to AMD features, the same is true for the other way around. AMD also offers a lot of new features via open-source instead of using proprietary closed designs.
For example, PureHair was an open-source, while HairWorks was not. Moreover, FidelityFX/ Contrast Aware Sharpening (CAS) is an open-source that works with all GPUs. On the other hand, Ansel and various NVIDIA features are available for NVIDIA GPUs only.
And if we talk about the features that were introduced five years ago like AMD’s Mantle API, they eventually helped to spur DirectX 12 and Vulkan features.
Besides features and technology, the manufacturing process also plays a big role in the performance of the current GPUs. AMD invested in TSMC’s 7nm FinFET node last year and it paid off well by helping the company’s Zen 2 architecture-based Ryzen CPUs surpass Intel’s CPUs in many aspects. Moreover, it also contributed to gains in efficiency on the Navi GPU front. That’s where AMD earns some points for being more aggressive in adopting newer manufacturing nodes.
Now, both NVIDIA and AMD are expected to have new chips using 7nm by the end of this year and the competition among the two for manufacturing the best GPU would heat up further.
So, when we look at the next-generation hardware for the GPUs, NVIDIA takes the lead for ray-tracing and other such enhancements to Vulkan and DirectX. Ray-tracing is a long-term play for sure and it could likely become a standard approach to graphics rendering in games over the next decade or so. Other features like Tensor cores for real-time de-noising and DLSS can open the door for installing ray-tracing in the GPUs with lower specifications.
Drivers and Software
When it comes to drivers and software, it’s difficult to determine a clear winner between AMD and NVIDIA. Quite a few users reported black screen issues with AMD drivers on RX 5000 Navi series GPUs, while others didn’t face any such issues. Newer drivers are supposed to fix these problems, but some users have still complaints. In this regard, NVIDIA drivers are also not foolproof. Because issues crop up for both companies depending on the game being played and the hardware used.
AMD always makes a lot of noise about its yearly driver overhaul. Recently, they aimed to simplify things by consolidating everything under one large umbrella with the Radeon Adrenalin 2020 drivers. Moreover, AMD pushed out seven driver updates this year out of which two (20.2.2 and 20.1.3) are Windows Hardware Quality Labs (WHQL) certified by Microsoft. This means that you can generally count on at least one new AMD driver per month with a minimum level of functionality requirement unless there are some major game launches.
On the other hand, NVIDIA’s driver schedule follows a similar pattern; you’ll get new drivers for major game launches or new graphics card hardware.
So far, six major driver releases have been released this year that are all WHQL certified. Furthermore, four hotfix driver updates have also been released. One of the major differences between AMD and NVIDIA drivers is that NVIDIA uses two separate user interfaces. The NVIDIA Control Panel also handles things like resolutions and certain graphics settings, while GeForce Experience tackles game optimizations, driver updates, and additional features like ShadowPlay, Ansel, and more.
However, you need to log in and solve a captcha prompt to use GeForce Experience every time, which might be annoying sometimes.
In terms of drivers and software, both AMD and NVIDIA show similar performance. Most of the users would prefer AMD’s unified driver approach since it’s one less interface to navigate. AMD’s drivers have also improved substantially since the days of the Catalyst Control Center, but sometimes less is more.
To determine which GPUs offer the better value between AMD and NVIDIA, we have broken down the pricing by looking at the current GPUs from both companies.
AMD may not have the fastest or most power-efficient GPUs as discussed above, but it can often sell you competitive performance at a lower price or the least that used to be the case.
When comparing the extreme end of the pricing scale, we can see that AMD doesn’t have to compete. Just forget the Titan RTX flagship at $2,500 or even the RTX 2080 Ti flagship that carries a price tag of $1,200. The RTX 2070 Super and RTX 2080 Super by AMD also cost substantially more than anything. So, we can comfortably say that AMD currently sells that’s worth buying.
However, if you are looking for the fastest GPU around, NVIDIA will serve the purpose for sure. Technically, NVIDIA is still giving the best value in extreme GPUs.
While comparing the high-end market starting from $350 to $500, AMD’s most-expensive GPU that you should even consider buying right now is the RX 5700 XT that costs $400. Before the impact of COVID-19, the prices of the GPUs were dipping as low as $350.
And if you’re willing to deal with mail-in rebates, you can still nab an MSI 5700 XT or ASRock 5700 XT for $360. Without the rebate, it will cost you $380. However, you won’t get support for ray-tracing in AMD GPUs.
When we compare further, though AMD’s 5700 XT flagship is about 13% faster than NVIDIA’s RTX 2060 Super on average, the latter comes with price tags starting from $400. Alternatively, the RTX 2070 Super is about 5% faster than the 5700 XT, but the prices jump to $500. Thus, AMD clearly wins over the value proposition for the high-end sector.
When coming down to the mid-range sector that costs around $230 to $350, we have five solid propositions with a pretty wide gap in performance and features. In this range, the RX 5700 is about 22% faster than the RX 5600 XT with 14 Gbps memory, which in turn leads NVIDIA’s RTX 2060 in performance by 5%.
On the other way around, the RTX 2060 is around 17% faster than the GTX 1660 Ti, which is only 1-2% faster than the GTX 1660 Super. Ultimately, NVIDIA wins when the RX 5600 XT flagship is compared with the RTX 2060 flagship because of the features since performance is basically tied.
However, we’ll recommend the RX 5700 GPU if your budget is around $300. If you only want to spend around $250 then we have RX 5600 XT and GTX 1660 Ti from AMD and NVIDIA respectively. And it’s obvious that AMD wins here because of the performance and features added in the RX 5600 flagship. So, AMD grabs the position in the mid-range value category.
And when budget GPUs are compared that are priced below $230 like GTX 1050 from NVIDIA and RX 560 from AMD, NVIDIA shows a better performance. Although RX 570 4GB cards are available at $120, the new RX 5500 XT 4GB costs 33% more and gives about 25% faster performance. Moreover, the latter uses half as much power.
However, we have GTX 1650 Super from NVIDIA which offers the same performance as the RX 5500 XT 4GB with higher power efficiency. According to our testing, the GTX 1650 Super consumes 25W less power than the RX 5500 XT 4GB. Thus, NVIDIA wins here over AMD.
When we look further into the budget spectrum, things get messy for the GPUs by the companies. Here, NVIDIA’s GTX 1660 Super is 20% faster than AMD’s RX 5500 XT 8GB but the former costs 28% more than the latter.
On the other hand, the vanilla GTX 1660 GDDR5 is only 3% faster, but it costs about 10% more, so AMD’s RX 5500 XT 8GB has comparatively better value. Overall, there’s a lot of flux in pricing on the budget GPU range, so we’re going to declare the budget segment a tie.
For the ultra-budget range, NVIDIA takes the lead over AMD with its GTX 1650, 1650 GDDR6, and 1650 Super flagships. While, AMD has an edge if you’re willing to spend a bit more money.
After looking at the whole pricing spectrum, AMD ends up being the value leader for sure. Although there’s a lot of subjectivity here, we give more weight to the mid-range and high-end price categories since that’s where a lot of our favorite GPUs reside. Although RTX 2080 Super and RTX 2080 Ti show great performance, most of the gamers are far more likely to buy a $250-$350 graphics card.
By winning over four categories out of six, NVIDIA continues to lead the GPU manufacturing market. Although AMD lands some solid blows against NVIDIA, particularly as a value proposition, we’re definitely interested in seeing how Ampere and Big Navi stir things up later this year.
As the bottom line, we’d say if you’re looking at the big picture, including performance, power efficiency, features, and the underlying technology, NVIDIA is in the lead. Even NVIDIA could win the drivers and software categories for showing better performance.
However, if you’re looking for the best graphics card within your budget, then the AMD-powered graphics card could be your choice. Here, NVIDIA won’t be of any help to you when we see the current prices of the highest-performance graphics cards. Since it has no real competitor for the extreme performance category for more than five years.
Hopefully, the RTX 20-series launch this year could put some pressure on NVIDIA to cut the prices for the high-end range. Moreover, a 50% improvement in performance per watt with RDNA 2 could close the gap between the GPUs from the two companies.
At the moment, NVIDIA continues to dominate. There are some aspects of GPUs that we didn’t delve into, like supercomputers, data centers, and deep learning hardware, but NVIDIA wins in these categories as well. Moreover, NVIDIA Quadro cards are already popular among most of the users. AMD may have taken the lead in the CPU race over Intel, but it still needs to compete with NVIDIA in the GPU war. Perhaps, its profits on the processor side will give the company the strength it needs to gain ground and overtake NVIDIA at some point.
Regarding the upcoming flagships from both AMD and NVIDIA, competition is important to drive performance and features forward while bringing the GPU prices down to affordable levels.