Best Graphics Card To Buy Right Now

You Can Trust Our Reviews

Since 1982, PCMag has tested and rated thousands of products to help you make ameliorate ownership decisions. (See how we test.)

Editor's Note: Before you dive into this guide, equally you'll come across from the Amazon prices above, the availability and pricing situation for GPUs is anything but "normal" correct now. If you lot plan to purchase a card soon, also meet this buying-strategies guide for communication on finding cards at a fair price. If you lot want to wait it out a bit longer, cheque out this how-to tutorial on getting the most performance from the GPU you already own.

Too annotation: Our picks above are based (in ascending lodge) on your target gameplay resolution, with picks for the virtually advisable Nvidia and AMD cards for each usage scenario (unless 1 is an unequivocal clear choice). We've factored in just a sampling of tertiary-political party cards here; many more than make full out the market. You can have our recommendation of a single reference carte in a given card form (such equally the GeForce GTX 1660 Ti, or the Radeon RX 6700 XT) equally an endorsement of the GPU family equally a whole.

If you're a PC gamer, or a content creator who lives and dies by the speed of your graphics-accelerated software, your video card is the engine that powers what you tin practise—or how lustily yous can brag.

Our guide will help you lot sort through the best video-menu options for your desktop PC, what you demand to know to upgrade a system, and how to evaluate whether a particular card is a adept buy. Nosotros'll too touch on some upcoming trends—they could affect which carte du jour yous choose. After all, consumer video cards range from under $100 to well over $one,499 (and that'southward only MSRP...more than on that after). It'due south easy to overpay or underbuy...but we won't let yous exercise that.

Who's Who in GPUs: AMD vs. Nvidia

First off, what does a graphics carte du jour practice? And practice you really need one?

If you're looking at any given prebuilt desktop PC on the market place, unless it's a gaming-oriented automobile, PC makers will de-emphasize the graphics card in favor of promoting CPU, RAM, or storage options. Indeed, sometimes that's for good reason; a depression-cost PC may not have a graphics card at all, relying instead on the graphics-accelerated silicon congenital into its CPU (an "integrated graphics processor," commonly called an "IGP"). In that location's nothing inherently wrong with relying on an IGP—most business laptops, cheap consumer laptops, and budget-minded desktops accept them—just if you're a gamer or a creator, the right graphics carte du jour is crucial.

(Photograph: Zlata Ivleva)

A mod graphics solution, whether information technology's a detached video carte du jour or an IGP, handles the display of second and 3D content, drawing the desktop, and decoding and encoding video content in programs and games. All of the detached video cards on the consumer market place are built around large graphics processing fries designed by ane of two companies: AMD or Nvidia. These processors are referred to as "GPUs," for "graphics processing units," a term that is also applied, confusingly, to the graphics card itself. (Nothing about graphics cards...ahem, GPUs...is simple!)

(Photograph: Zlata Ivleva)

The two companies work up what are known as "reference designs" for their video cards, a standardized version of a bill of fare congenital effectually a given GPU. Sometimes these reference-design cards are sold directly past Nvidia (or, less often, by AMD).

Nvidia'southward own make of cards are spotted easily by their "Founders Edition" branding, something that until the release of Nvidia's latest, the GeForce RTX 3000 series, didn't mean much more than slightly higher clock speeds from stock and sturdy build quality. Often the Founders Editions cards are the most aesthetically consequent of any cards that might come out during the lifetime of a particular GPU. But their designs tend to be conservative, non as accommodating to aggressive overclocking or modification every bit some third-political party options are.

Nvidia's new designs for its Founders Edition cards throw nigh of the conventional wisdom out the window. The cards are packed onto a PCB ("printed excursion lath," the guts of a GPU), that is l% smaller than previous-generation RTX 20 Series Nvidia GPUs in each corresponding model that features the visitor's new "push-pull" cooling organisation. Nvidia's engineering science talent has been on full display in these cards, and although AMD puts up a expert fight on performance, if advanced industrial pattern is your thing, the RTX thirty Series Founders Edition cards stand alone.

Nvidia GeForce RTX 3060 Ti Founders Edition

(Photo: Zlata Ivleva)

This makes Nvidia's Founders Edition cards smaller, lighter, and faster than e'er before, but and so far we've only seen cards carrying Nvidia's Founders Edition badge with this treatment. Sometimes reference cards are duplicated by third-party menu makers (companies referred to in industry lingo every bit AMD or Nvidia "board partners"), such as Asus, EVGA, MSI, Gigabyte, Sapphire, XFX, and Zotac. Depending on the graphics chip in question, these board partners may sell their own cocky-branded versions of the reference card (adhering to the design and specifications set past AMD or Nvidia), or they will way their own custom products, with different cooling-fan designs, slight overclocking done from the factory, or features such as LED mood illumination. Some board partners will do both—that is, sell reference versions of a given GPU, as well as their ain, more radical designs.

Tertiary-political party cards similar the MSI GeForce RTX 3080 Gaming X Trio 10G are virtually as classic as GPU design gets, and AMD'southward competing Radeon RX 6000 Series, while slick in their ain correct, don't feature the kind of technological frog leaps in cooling and power efficiency we've seen on display in electric current-gen Founders Edition cards.

Who Needs a Discrete GPU?

Nosotros mentioned integrated graphics (IGPs) above. IGPs are capable of coming together the needs of most general users today, with 3 broad exceptions...

Professional Workstation Users. These folks, who work with CAD software or in video and photo editing, will notwithstanding benefit profoundly from a detached GPU. Some of their cardinal applications can transcode video from i format to another, or perform other specialized operations using resources from the GPU instead of (or in improver to) those of the CPU. Whether this is faster will depend on the awarding in question, which specific GPU and CPU yous ain, and other factors.

Productivity-Minded Users With Multiple Displays.People who demand a large number of displays can also benefit from a discrete GPU. Desktop operating systems can drive displays continued to the IGP and discrete GPUs simultaneously. If you've ever wanted v or six displays hooked up to a unmarried organization, you can combine an IGP and a discrete GPU to get there.

That said, you lot don't necessarily need a high-cease graphics menu to do that. If y'all're simply displaying concern applications, multiple browser windows, or lots of static windows across multiple displays (i.e., not enervating PC games), all you demand is a card that supports the brandish specifications, resolutions, monitor interfaces, and number of panels you demand. If you're showing 3 web browsers across three display panels, a GeForce RTX 3080 card, say, won't confer any greater benefit than a GeForce GTX 1660 with the aforementioned supported outputs.

Gamers.And of grade, there's the gaming market, to whom the GPU is arguably the most important component. RAM and CPU choices both matter, but if you take to selection between a top-end system circa 2018 with a 2022 GPU or a top-cease system today using the highest-cease GPU you could purchase in 2018, yous'd desire the former.

Graphics cards fall into two distinct classes: consumer cards meant for gaming and light content creation work, and dedicated cards meant for professional workstations and geared toward scientific computing, calculations, and artificial intelligence piece of work. This guide, and our reviews, will focus on the former, but nosotros'll touch on workstation cards a little chip, later on on. The key sub-brands you demand to know across these ii fields are Nvidia'southward GeForce and AMD's Radeon RX (on the consumer side of things), and Nvidia'due south Titan and Quadro (now RTX A-Series), as well as AMD's Radeon Pro and Radeon Instinct (in the pro workstation field). Nvidia continues to dominate the very high end of both markets.

(Photo: Zlata Ivleva)

For at present though, we'll focus on the consumer cards. Nvidia's consumer carte du jour line in early 2022 is cleaved into ii distinct classes, both united under the long-running GeForce brand: GeForce GTX, and GeForce RTX. AMD'due south consumer cards, meanwhile, comprise the Radeon RX and (now by and large gone) Radeon RX Vega families, likewise as the end-of-life Radeon Seven. Before nosotros get into the private lines in detail, though, allow'south outline a few very important considerations yous should make for any video-carte du jour buy.

Target Resolution and Monitor Tech: Your First Considerations

Resolution is the horizontal-by-vertical pixel count at which your video carte du jour will drive your monitor. This has a huge bearing on which bill of fare to buy, and how much y'all need to spend, when looking at a video card from a gaming perspective.

If you lot are a PC gamer, a big part of what yous'll want to consider is the resolution(southward) at which a given video card is best suited for gaming. Nowadays, even low-terminate cards will display everyday programs at lofty resolutions similar 3,840 by 2,160 pixels (a.k.a., 4K). Just for strenuous PC games, those cards will not have virtually the ability to drive shine frame rates at loftier resolutions like those. In games, the video menu is what calculates positions, geometry, and lighting, and renders the onscreen paradigm in real time. For that, the higher the in-game item level and monitor resolution you lot're running, the more graphics-menu muscle is required.

Resolution Is a Central Decision Point

The three about common resolutions at which today's gamers play are 1080p (1,920 by 1,080 pixels), 1440p (two,560 by one,440 pixels), and 2160p or 4K (3,840 by 2,160 pixels). More often than not speaking, you'll desire to choose a carte du jour suited for your monitor'south native resolution. (The "native" resolution is the highest supported by the panel, and the ane at which the display looks the best.)

You'll also see ultra-wide-screen monitors with in-betwixt resolutions (3,440 by i,440 pixels is a common ane); you can gauge these versus 1080p, 1440p, and 2160p by computing the raw number of pixels for each (multiply the vertical number by the horizontal one) and seeing where that screen resolution fits in relative to the mutual ones. (See our targeted roundups of the best graphics cards for 1080p play and the all-time graphics cards for 4K gaming.)

Why does this matter? Well, in the case of PC gaming, the ability of the components within your next PC—whether yous are buying i, edifice ane, or upgrading—should be distributed in a style that all-time suits the manner you want to play.

(Photo: Zlata Ivleva)

Without getting too deep into the weeds, hither'due south how it works: The frame rates you'll come across when gaming at 1080p, even at the highest detail levels, are most e'er down to some balance of CPU and GPU power, rather than either 1 existence the outright determinant of pinnacle frame rates.

Next is the 1440p resolution, which starts to dissever the load when you are playing at higher detail levels. Some games start to inquire more than of the GPU, while others tin still lean on the CPU for the heavy math. (It depends on how the game has been optimized by the developer.) Then there's 4K resolution, where, in most cases, almost all of the lifting is washed exclusively by the GPU.

Now, of grade, you can ever dial down the detail levels for a game to make it run acceptably at a higher-than-recommended resolution, or dial dorsum the resolution itself. Only to an extent, that defeats the purpose of a graphics carte purchase. The highest-stop cards are meant for 4K play or for playing at very high refresh rates at 1080p or 1440p; you don't have to spend $ane,000 or even one-half that to play more than acceptably at 1080p.

In short: E'er buy the GPU that fits the monitor you either play on today or programme to own in the near future. In that location are plenty of midrange GPUs that can power 1440p displays at their peak, and 4K is still, at present, a fringe display resolution for the most agile PC gamers if the Steam Hardware Survey(Opens in a new window) is any indication. (It saw less than 3% of users playing at 4K in early 2022.)

High-Refresh Gaming: Why High-End GPUs Affair

Another thing to go along abreast of is a trend in gaming that's gained major momentum in recent years: loftier-refresh gaming monitors. For ages, 60Hz (or sixty screen redraws a second) was the panel-refresh limit for most PC monitors, but that was before the genre of esports really hit its pace.

Panels focused on esports and loftier-refresh gaming may back up upwardly to 144Hz, 240Hz, or even 360Hz for smoother gameplay. What this means: If y'all have a video card that can consistently button frames in a given game in excess of 60fps, on a high-refresh monitor you may exist able to encounter those formerly "wasted" frames in the course of smoother game motion.

Powered by esports success stories (like 16-year-one-time Fortnite prodigy Bugha(Opens in a new window) turning into a multi-millionaire overnight), the need has surged in recent years for loftier-refresh monitors that tin continue esports hopefuls playing at their peak. And while 1080p is withal overwhelmingly the preferred resolution for competitive players across all game genres, many are post-obit the trends that monitors are setting first.

(Photo: Zlata Ivleva)

The number of players moving upward to the 1440p subclass of graphical resolutions (played in either 16:nine aspect ratio at 2,560 by 1,440 pixels, or in 21:nine at iii,440 by ane,440) is growing faster than always, thanks in no pocket-sized part to contempo game-monitor entries like the ViewSonic Elite XG270QG, which marries the worlds of loftier-refresh and high-quality panels. To an extent, the cards and the panels are playing a game of leapfrog themselves.

Gaming at a college resolution does have its benefits for those who desire to hit their opponents with pixel-perfect precision, simply but every bit many esports hopefuls and currently salaried pros still swear by playing at resolutions as low as 720p in games like Counter-Strike: Global Offensive. So all told, your mileage may vary, depending on the way y'all prefer to play, every bit well as on which games you play.

Nearly coincidental gamers won't care about extreme refresh rates, but the difference is marked if you lot play fast-action titles, and competitive esports hounds detect the fluidity of a high refresh charge per unit a competitive advantage. (See our picks for the best gaming monitors, including high-refresh models.) In brusk: Buying a powerful video card that pushes high frame rates can be a boon nowadays even for play at a "pedestrian" resolution like 1080p, if paired with a high-refresh monitor.

HDR Compatibility

Finally, continue HDR compatibility in mind. More and more monitors these days—including about every one of our Editors' Choice picks for best gaming monitor of late—support HDR at some level. And while in our testing HDR 10 and HDR 400 monitors don't often make much impact for their HDR epitome quality, any monitors in a higher place the HDR 600 spec should cistron into your GPU decision every bit both a display for gaming and one for HDR-enhanced content.

(Photo: Zlata Ivleva)

Monitor buyers should likewise make sure the model they choose supports HDR transfer at a refresh rate and bitrate that a new card can support. It'due south a dance, but i that can pay off beautifully on content creation and gaming monitors all the aforementioned.

FreeSync vs. G-Sync: Jets! Sharks! Maria?

Should you buy a carte du jour based on whether it supports 1 of these ii venerable specs for smoothing gameplay? It depends on the monitor you have.

FreeSync (AMD's solution) and G-Sync (Nvidia's) are two sides of the same coin, a engineering called adaptive sync. With adaptive sync, the monitor displays at a variable refresh rate led past the video carte; the screen draws at a rate that scales up and downwardly according to the card'southward output capabilities at any given time in a game. Without information technology, wobbles in the frame rate can pb to artifacts, staggering/stuttering of the onscreen action, or screen tearing, in which mismatched screen halves display momentarily. Under adaptive sync, the monitor draws a full frame only when the video bill of fare can deliver a whole frame.

(Photo: Chris Stobing)

The monitor y'all own may back up FreeSync or G-Sync, or neither. FreeSync is much more than common, equally information technology doesn't add together to a monitor's manufacturing cost; G-Sync requires defended hardware within the display. You may wish to opt for i GPU maker's wares or the other's based on this, only know that the tides are changing on this front end. At CES 2019, Nvidia announced a commuter tweak(Opens in a new window) that allows FreeSync-compatible monitors to use adaptive sync with late-model Nvidia GeForce cards, and a rapidly growing subset of FreeSync monitors has been certified(Opens in a new window) by Nvidia as "Grand-Sync Compatible." So the choice may not be as black and white (or as reddish or green) as it has been for years.

Nosotros've tested both, and unless you're competing in a CS:GO or Overwatch pro-am excursion, you might be hard-pressed to see whatever consequent difference between the two in the latest models. Screen vehement was a more difficult problem to solve back when 1000-Sync was first introduced, and these days both FreeSync and Chiliad-Sync-Uniform monitors work well plenty that simply expert optics tin tell the difference.

Encounter the Radeon and GeForce Families

Now that nosotros've discussed the ways these ii rival gangs take come together in recent years, now let's talk about what makes them dissimilar. The GPU lines of the two big graphics-chip makers are constantly evolving, with depression-terminate models suited to low-resolution gameplay ranging up to elite-priced models for gaming at 4K and/or very loftier refresh rates. Let's look at Nvidia'southward first. (Again: Bear in heed the volatility of today's prices versus the listing prices we mention, and know that MSRPs are only that these days: "suggested" prices!)

A Wait at Nvidia'due south Lineup

Until the terminate of 2020, the primary part of the company's carte du jour stack was split between cards using concluding-generation (a.g.a. "20-serial") GPUs dubbed the "Turing" line, and newer GTX 1600 series cards, also based on Turing architecture. The very newest introductions, the GeForce RTX thirty-Serial cards, are high-end cards based on GPUs using an architecture called "Ampere."

Nvidia GeForce RTX 3070 Founders Edition Card

(Photo: Zlata Ivleva)

Here's a quick rundown of the currently relevant card classes in the "Pascal" (Turing'south predecessor), Turing, and Ampere families, their rough pricing, and their usage cases...

If you're a longtime observer of the market, you'll detect that many of the aging GeForce GTX Pascal cards like the GTX 1070 and GTX 1080 are not listed in a higher place. They have sold through and are largely found on the second-hand marketplace in 2022 in favor of their GeForce RTX successors. The GeForce GTX 1060 has met a similar fate due to the release of the GeForce GTX 1660 and 1660 Ti, while the GTX 1050 has since fallen out of relevance to the GTX 1650 and GTX 1650 Super.

But start, let's talk Turing. When Nvidia launched its line of 20 Series GPUs in September of 2018, the reaction was mixed. On the one mitt, the company was offer up some of the nigh powerful GPUs seen to date, consummate with new and exciting technologies like ray-tracing and DLSS. But on the other, at the time of the Turing launch, no games supported ray-tracing or DLSS. Even two years later, the library of titles that supports DLSS ii.0 on its own or combined with ray-tracing is limited.

(Photo: Zlata Ivleva)

At the same fourth dimension, Nvidia also moved the goalposts for high-end GPU pricing, compared with by generations. The GeForce RTX 2080 Ti, the company's new flagship graphics card, would hit shelves in excess of $1,000, and the side by side menu downward, the $699 GeForce RTX 2080, wasn't much better.

(Photo: Zlata Ivleva)

The visitor course-corrected in 2019, releasing the GeForce RTX 2060 Super, RTX 2070 Super, and RTX 2080 Super (upwardly-ticked versions of the existing cards) at the aforementioned time that AMD was launching its AMD Radeon RX 5700 and RX 5700 XT midrange GPUs. Covering both the RTX and GTX segments, Nvidia's Super cards heave the specs of each card they're meant to supervene upon in the stack (some more effectively than others).

This all brings us to September 2020, and the launch of the GeForce RTX 30 Serial. Nvidia unveiled new GeForce RTX 3070, GeForce RTX 3080, and GeForce RTX 3090 GPUs. An RTX 3080 Ti, RTX 3070 Ti, RTX 3060 Ti, RTX 3060, and RTX 3050 came later. They are a big enough deal to merit their own spec breakout. (Note: At that place is no Founders Edition of the RTX 3060 and 3050, only third-party models.)

The cards, built off Samsung'southward 8nm process, are a generational leap, moving RT cores to their second generation, Tensor cores to their tertiary, and the retention type from GDDR6 to GDDR6X. Reworked PCBs accept mandated tons of new innovations in everything from the placement of various modules and chips on the PCB to the inner workings of a brand-new heatsink.

As far equally how the 30 Series has affected costs up and downward the Nvidia card stack, we'd still classify the GeForce GT 1030 to GTX 1050 as very low-finish cards, at under $100 or a petty above. The GTX 1650/GTX 1650 Super to GTX 1660 Ti, RTX 2060 Super and RTX 3060 make upwards Nvidia'south electric current low-to-midrange, spanning almost $150 to $300 or a little higher at listing price.

(Photo: Zlata Ivleva)

The midrange and high end got a whole lot more complicated with the release of the GeForce RTX 30 Serial. We'd put the GeForce RTX 3080 and RTX 3090 in an "aristocracy" loftier-end pricing category, starting at $699 (once again; MSRP!) and going upward from at that place, split up from the more midrange (but withal enough powerful) options similar the RTX 2060, RTX 2070, RTX 2080, RTX 3060 Ti, and RTX 3070. These cards will generally start around $350 MSRP, and in the current pricing structure, will range up to about $650 at MSRP depending on the model.

A Look at AMD'south Lineup

Equally for AMD'southward carte du jour classes, in 2022 the company is stronger than it has been for some fourth dimension, competing ably enough with Nvidia's low-end, mainstream, and high-finish cards.

The aging Radeon RX 550 and RX 560 incorporate the depression finish, while the old Radeon RX 570 to RX 590 are the former midrange and platonic for 1080p gaming, though their time is limited, given the visitor's latest additions to its 1080p-play armory, the Radeon RX 5500 XT and the Radeon RX 5600 XT, plus the even newer Radeon RX 6600 and RX 6500 XT. The Radeon RX 580, RX Vega 56, and RX Vega 64 cards, the first is even so a peachy-value 1080p carte du jour and the latter two good for both 1080p and 1440p play, are decidedly yesterday's cards.

(Photo: Zlata Ivleva)

Indeed, the 1080p and especially the 1440p AMD cards have seen a shakeup. The company released the starting time of its new, long-awaited line of 7nm-based "Navi" midrange graphics cards in July of 2019 based on a whole new compages, Radeon DNA (RDNA). The first three cards were the Radeon RX 5700, the Radeon RX 5700 XT, and the limited-run Radeon RX 5700 XT Ceremony Edition. All these cards accept their sights pinned on the 1440p gaming market. Each, indeed, powers demanding AAA titles at above 60fps in that resolution subclass.

To boxing Nvidia at the top-stop, AMD'southward Radeon RX 6700 XT, RX 6800, RX 6800 XT, and RX 6900 XT took up the mantle in the beginning half of 2021. These RDNA 2-based cards offer amend price-to-functioning than the company's previous-generation RDNA 1 cards; however, they're all the same by and large a few percentage points behind Nvidia's thirty Series Founders Edition cards in design, cooling, and driver stability.

In 2020, the visitor pulled back the mantle on RDNA 2. Featured in both the company's discrete desktop graphics cards as well equally the Sony PS5 and Xbox Series Ten, RDNA 2 refines many of the elements that debuted in RDNA, and while simultaneously expanding a new fix of features that aim to keep AMD neck and neck with Team Greenish. This includes ray-tracing compute cores, and support for Microsoft's DX12 Ultimate API. Here's a look at the first few RDNA two cards...

We've found in our fourth dimension with both the new AMD Radeon RX 6800 XT and the Radeon RX 6800 "Big Navi" high-end cards, both cards were able to keep footstep with Nvidia's biggest and burliest brawlers, similar the GeForce RTX 3080 Founders Edition. All the same, during our benchmarking, a not-unique problem reared its caput once again: inconsistent frame rates with older games, at times, and the occasional commuter stability. On some games, we saw slower-than-expected frame rates, and in a few spots, graphical glitches that break the engines of games like PUBG.

(Photo: Zlata Ivleva)

AMD has said it'south enlightened of the issues with its launch drivers and nosotros'll be investigating the bug as time goes on, but as information technology stands at this writing, Nvidia'due south Ampere cards have provided a more stable gaming experience throughout each of the benchmarks we run as a part of our testing suite.

Graphics Carte du jour Nuts: Agreement the Core Specs

Now, our comparison charts above should give you a good idea of which card families you should exist looking at, based on your monitor and your target resolution (and your budget). A few fundamental numbers are worth keeping in mind when comparing cards, though: the graphics processor's clock speed, the onboard VRAM (that is, how much video memory it has), and—of course!—the pricing.

Clock Speed

When comparison GPUs from the same family unit, a higher base clock speed (that is, the speed at which the graphics core works) and more cores signify a faster GPU. Once more, though: That'due south only a valid comparing between cards in the same product family based on the same GPU. For instance, the base clock on the Founders Edition GeForce GTX 3080 is 1,710MHz, while the base clock is 1,815MHz on a (factory overclocked) Gaming X Trio version of the RTX 3080 (using the same flake) from MSI in its out-of-the-box Gaming Mode.

(Photo: Zlata Ivleva)

Annotation that this base clock measure is distinct from the graphics bit's boost clock. The boost clock is the speed to which the graphics chip can accelerate temporarily when under load, equally thermal conditions let. This tin can also vary from card to card in the same family. Information technology depends on the robustness of the cooling hardware on the card and the aggressiveness of the manufacturer in its factory settings. The peak-end partner cards with behemothic multi-fan coolers will tend to have the highest heave clocks for a given GPU.

This is to say null of AMD'southward proprietary category of GPU speed: "game clock." According to the visitor, game clock represents the "average clock speed gamers should expect to see across a broad range of titles," a number that the company's engineers gathered during a test of 25 different titles on the company's RDNA and RDNA 2-based lineup of cards. Nosotros mention this so that you don't compare game clocks to boost or base clocks, which game clock incomparably is not.

Agreement Onboard Video-Bill of fare Memory

The corporeality of onboard video retention (sometimes referred to by the erstwhile-school term "frame buffer") is usually matched to the requirements of the games or programs that the carte du jour is designed to run. In a certain sense, from a PC-gaming perspective, you can count on a video carte du jour to take enough retentiveness to handle current enervating games at the resolutions and detail levels that the card is suited for. In other words, a carte maker generally won't overprovision a bill of fare with more retentiveness than information technology can realistically use; that would inflate the pricing and make the carte less competitive. But there are some wrinkles to this.

A carte du jour designed for gameplay at 1,920 by ane,080 pixels (1080p) these days will generally be outfitted with 4GB or 6GB of RAM, while cards geared more toward play at 2,560 by one,440 pixels (1440p) or iii,840 by 2,160 (2160p, or 4K) tend to deploy 8GB or more. Usually, for cards based on a given GPU, all of the cards take a standard amount of memory.

The wrinkles: In some isolated but important cases, bill of fare makers offer versions of a card with the same GPU simply unlike amounts of VRAM. A proficient example are cards based on the Radeon RX 5500 XT, which comes in both 4GB and 8GB variants. Both are GPUs you'll discover in popular midrange cards, then heed the retention amount on these. The cheaper versions will have less.

(Photo: Zlata Ivleva)

At present, if you're looking for a video card for all-out 1080p gameplay, a menu with at to the lowest degree 4GB of memory really shouldn't exist negotiable. Both AMD and Nvidia now outfit their $200-MSRP-plus GPUs with more VRAM than this. (AMD has stepped up to 8GB on its RX-series cards, with 16GB on its pinnacle-end ones, while Nvidia is using 6GB or 8GB on virtually, with 24GB on its elite GeForce RTX 3090.) Either way, sub-4GB cards should only be used for secondary systems, gaming at low resolutions, or elementary or older games that don't need much in the way of hardware resource.

For creators, it's an entirely different ballgame. In many 3D rendering programs (as well as VFX workflows, modeling, and video editing), specs like the boost clock speed are about less important of a determination point than the amount of onboard VRAM. The more VRAM, the faster the memory, and the larger the bandwidth pipage a bill of fare has, the amend it will be (in most cases) for a task like rendering out a complex VFX scene that has thousands, if not millions, of different elements to calculate at in one case.

Retentiveness bandwidth is another spec you will see. Information technology refers to how quickly information tin can move into and out of the GPU. More than is generally better, but again, AMD and Nvidia have unlike architectures and sometimes different memory bandwidth requirements, and then these numbers are non directly comparable.

Retention type is too an of import factor of your adjacent GPU purchase, and knowing which type you're ownership into is critical depending on the types of games you play or programs yous program to run, though you won't really have a pick within a carte du jour line.

High-Bandwidth Memory 2 (HBM2): AMD went all-in on HBM2 with the release of the AMD Radeon Seven, a card that tin technically keep up with an RTX 2080 in gaming, but too packs in a whopping 16GB of HBM2 VRAM for content creators. HBM2 is actually preferred for a particular subset of workloads like video editing in Adobe Premiere Pro, and was favored by AMD for its cheaper manufacturing cost. It's since fallen out of favor to GDDR6.

GDDR6: Considered the workhorse of the modern GPU, GDDR is a memory type that lives in almost every menu released in the past few decades, from the RTX 2080 Ti and RTX Titan all the way downwards to the AMD Radeon RX 5600 XT. The latest version, GDDR6, is a reliable, highly tunable VRAM solution that fits every SKU of the market, and often provides more than plenty horsepower for the price to handle fifty-fifty the well-nigh demanding AAA games at high resolution. AMD opted for GDDR6 in its newest Radeon RX 6000 Series cards, which wouldn't seem like a bad upgrade if Nvidia'southward newest cards weren't already using...

GDDR6X: Nvidia has begun employing this new type of memory in its GeForce RTX 30 Serial. It finer doubles the available bandwidth of the original GDDR6 design, all without running into the same problems of signal degradation or path interference that previous iterations would need to account for.

Upgrading a Pre-Built Desktop With a New Graphics Carte du jour

Assuming the chassis is large enough, well-nigh pre-congenital desktops these days take enough cooling capability to have a new discrete GPU with no problems.

The outset thing to do earlier buying or upgrading a GPU is to measure your chassis for the available carte infinite. In some cases, yous've got a gulf between the far right-manus border of the motherboard and the difficult drive trophy. In others, you lot might accept barely an inch to spare on the total length of your GPU. Really long cards can present a problem in some smaller cases. (See our favorite graphics cards for compact PCs.)

Adjacent, check your graphics carte's tiptop. The card partners sometimes field their own menu coolers that depart from the standard AMD and Nvidia reference designs. Make certain that if your chosen carte du jour has an elaborate cooler pattern, it's non then tall that it keeps your case from closing.

MSI GeForce RTX Card Installed With RGB Lighting

(Photo: Zlata Ivleva)

Finally: the power supply unit (PSU). Your system needs to have a PSU that's upwardly to the task of giving a new carte plenty juice. This is something to be especially wary of if yous're putting a loftier-cease video card in a pre-built PC that was equipped with a low-end card, or no card at all. Doubly then if it'due south a budget-minded or business organization system; these PCs tend to accept underpowered or minimally provisioned PSUs.

The two most of import factors to exist aware of here are the number of vi-pin and eight-pin cables on your PSU, and the maximum wattage the PSU is rated for. Most modern systems, including those sold by OEMs like Dell, HP, and Lenovo, apply power supplies that include at least ane vi-pin power connector meant for a video card, and some have both a half dozen-pin and an viii-pin connector.

Midrange and loftier-stop graphics cards volition require a six-pin cable, an eight-pin cable, or some combination of the 2 to provide working power to the card. (The very lowest-end cards describe all the ability they need from the PCI Limited slot.) Brand sure y'all know what your bill of fare needs in terms of connectors.

Nvidia GeForce RTX Founders Edition Card Power Connectors

(Photo: Zlata Ivleva)

Nosotros've seen some changes here of belatedly, every bit some of the GeForce RTX Founders Edition cards require a special adapter (information technology comes in the box) to turn two eight-pivot PSU connectors into a single 12-pin 1 card-side, and the massive 12.7-inch MSI GeForce RTX 3080 Gaming Ten Trio (and a few other high-end monsters) now require a whopping three eight-pin connectors to suck down its required juice. And it's non the only new RTX card to crave three PSU connectors.

(Photo: Zlata Ivleva)

Nvidia and AMD both outline recommended power supply wattage for each of their graphics-card families. Have these guidelines seriously, only they are merely guidelines, and they are generally bourgeois. If AMD or Nvidia says you demand at to the lowest degree a 500-watt PSU to run a given GPU, don't chance it with the 300-watter yous may take installed, merely know that you don't need an 800-watt PSU to guarantee plenty headroom, either. Third-political party versions of a given GPU may vary slightly from AMD's and Nvidia'south advice for their reference cards, then always check the power-supply recommendation details for the specific card y'all are looking at.

SLI, CrossFireX, and NVLink: Fading for Gamers

Over the past few generations, both AMD and Nvidia have been moving away from back up for dual-, tri-, or even quad-carte setups. These have traditionally been a not-then-cheap, but somewhat like shooting fish in a barrel, fashion to maximize performance. Merely the value suggestion (for PC gamers, specifically) just isn't there anymore.

(Photo: Zlata Ivleva)

In our testing at PC Labs in 2019 with twin RTX 2080 Ti cards, we plant that adding ii cards to the mix provided, well...mixed results, to put it mildly. Nearly games these days aren't written to leverage two or more than cards, and those that do don't see performance scale upwardly in parallel. (SLI and NVLink are Nvidia's twin-carte du jour technologies; CrossFireX is AMD's.) Some games really run worse; information technology's all down to engine optimization.

For content creation tasks, though, it's a different story. There'south a reason why the GeForce RTX 3090 is the but card in Nvidia'southward current lineup that supports whatever kind of NVLink card-pairing: Pro-level creators are the only ones who will get enough use out of it to become a satisfactory return on investment.

Lesser line? In well-nigh all cases nowadays, you'll be best served by buying the unmarried all-time card you can beget, rather than buying i lesser card now, planning to accept another join it alongside later.

Ports and Preferences: What Connections Should My Graphics Card Have?

Iii kinds of port are common on the rear edge of a current graphics card: DVI, HDMI, and DisplayPort. Some systems and monitors still use DVI, but it's the oldest of the 3 standards and no longer appears on high-end cards these days.

Most cards have several DisplayPorts (often 3) and one HDMI port. When it comes to HDMI versus DisplayPort, note some differences. First, if yous plan on using a 4K brandish, now or in the future, your card needs to at least support HDMI 2.0a or DisplayPort 1.2/1.2a. It's fine if the GPU supports annihilation above those labels, like HDMI 2.0b or DisplayPort ane.four, simply that's the minimum yous'll want for smooth 4K playback or gaming. (The latest-gen cards from both makers will be fine on this score.)

HDMI ii.1 is a new cablevision spec compatible with all of Nvidia's GeForce RTX xxx Serial cards, which ups the old bandwidth limits from 18Gbps (in HDMI 2.0) to 48Gbps (in HDMI two.1). The upgrade besides enables 8K resolution to display at a refresh rate up to 60Hz, with 4K supported upward to 120Hz.

(Photo: Zlata Ivleva)

Annotation that some of the cards from Nvidia's GeForce RTX Turing line (the 20 Series) use a port called VirtualLink. This port looks like (and can serve as) a USB Type-C port that also supports DisplayPort over USB-C. What the port was actually designed for, though: attaching futurity generations of virtual-reality (VR) headsets, providing power and bandwidth adequate to the needs of VR head-mounted displays (HMDs). Information technology's nice to have, but no VR hardware supports it yet, and it's not on the Founders Edition cards of the RTX thirty Series, then its future is cloudy.

Looking Forrard: Graphics Carte du jour Trends

Nvidia has been in the consumer video menu driver's seat for a few years now, but the last 18 months have seen more menu action than any similar period in recent retention, shaking things up between the 2 big players.

Image Sharpening Tech, Super Resolution, and Nvidia DLSS

A major alter to the mural of gaming over the last year or so has been the addition of image-sharpening technologies: Radeon Image Sharpening (RIS) by AMD, FidelityFX with CAS (as well by AMD), and Freestyle from Nvidia. But what are these programs, exactly, and how do they aid gamers who are shopping on a upkeep?

Information technology all has to practise with something called "render scaling." In almost mod games, you've probable seen something in your graphics settings that lets you alter the return scale of a game. In essence, what this does is take the electric current resolution you have the game set to (in this case, let's say information technology's 2,560 by 1,440 pixels) and pushes the "return" resolution down by a particular percentage, perhaps downwards to 2,048 by 1,152 pixels (again, for the sake of this example).

Simply wait...who would brand their game await worse on purpose? Users of game sharpeners, that'south who. Image-sharpening technologies let y'all scale down a game's render resolution, thereby increasing the frame rate (lower resolutions mean less pixels to draw for the GPU, and thereby less muscle needed), while a sharpener cleans things up on the back finish for a small operation price.

"Cleaning things upwards" involves applying a sharpening filter to the downsampled image. If you can tune information technology just right (85% down-render with a 35% acuminate scale is a popular ratio), in theory you can gain a pregnant amount of functioning with little discernible loss in visual clarity. Why is this of import? If you can render your game downwardly without losing visual quality, ultimately this ways you tin buy a bottom video card and become the same effective functioning.

Nosotros've pushed image-sharpening technologies to their limit, and in our testing establish that the meridian down-sample is most 30%. This means you lot can buy a card that'due south most a third cheaper than the one you were originally looking at, acuminate it back up 30% using ane of the same sharpening tools, and withal get shut to the same high-definition gaming experience yous would expect from running a game at its native resolution without render scaling.

On a related note, DLSS, short for "deep-learning supersampling," is Nvidia's new solution to a problem as old as 3D-capable video cards themselves: how to shine out the polygons around the edge of a character or object with equally picayune performance bear upon as possible. Anti-aliasing, as it's amend known, is i of the almost computationally difficult tasks for a graphics card to work through in video games, and since the technology'south inception, a wide assortment of approaches have used all mode of math to reach the same goal: make the jagged thing look smoother.

In the case of DLSS, Nvidia employs artificial intelligence to help. But for at present, DLSS comes at a premium (i.e., it requires a GeForce RTX card) because like ray-tracing, it can't exist done on just whatever ol' CUDA core: It has to happen on a specialized graphics core known as a Tensor cadre. What the RT core is to ray-tracing, the Tensor core is to decoding complex equations provided by Nvidia'southward artificial intelligence neural network.

DLSS and its follow-on, DLSS ii.0, show great hope; the main problem is that and then few games back up the tech. This is irresolute with the integration of DLSS ii.0 into Unreal Engine 4; however, it will however take some time earlier developers are using it on a broad or universal calibration. In our testing at PC Labs, nosotros institute that one of the highest-profile DLSS-capable games, Expiry Stranding, ever benefitted from the use of either DLSS or CAS. But in the case of DLSS specifically, the visual quality was really improved, whereas CAS got it perchance 90% of the mode there with some visible jitters that would appear while characters were in motion. If more games prefer DLSS, information technology could exist a huge boon to GeForce RTX carte owners. Simply "if" is the big word in that location.

AMD also threw some punches in 2021 in this fight, with its FidelityFX Super Resolution. While this upsampling technology was supported by only a handful of games at its mid-2021 launch, it uses an open-source arroyo to ease adoption, and more than than thirty games have been added to the listing of compatible titles since we tested the feature back in June 2021. It works with a much broader range of contempo video cards...from both Nvidia and AMD. This is unlike DLSS, which is tied to GeForce RTX cards.

Most recently, in the first half of 2022, AMD fired its newest weapon in the upscaling war: Radeon Super Resolution(Opens in a new window). Super Resolution works off the same principles and underlying algorithms that power FidelityFX Super Resolution, but they're now practical at the commuter level. This limits the GPUs it works on—just models in the AMD Radeon RX 5000 and RX 6000 family unit—only opens up the number of games that can support the feature into the tens of thousands. Check out the link for a full breakdown of how the tech works.

VR: New Interfaces, New HMDs?

Equally we alluded to with VirtualLink, VR is another consideration. VR's requirements are slightly different than those of unproblematic monitors. Both of the old mainstream tethered VR HMDs, the original HTC Vive and Oculus Rift, have an effective resolution across both optics of two,160 by 1,200. That's significantly lower than 4K, and it's the reason why older midrange GPUs like AMD's Radeon RX 5700 XT or Nvidia's GeForce GTX 1660 Super tin be used for VR.

On the other hand, VR demands higher frame rates than conventional gaming. Depression frame rates in VR (anything below 90 frames per 2d is considered low) can upshot in a bad VR gaming experience. Higher-end GPUs in the $300-MSRP-plus category are going to offer better VR experiences today and more than longevity overall, simply VR with current-generation headsets can be sustained on a lower-end carte du jour than 4K.

(Photo: Zlata Ivleva)

That said, in 2019, two new headsets upped the power requirements a fleck. The Oculus Rift S has raised the bar to a resolution of 2,560 by one,440 pixels per eye, while the mega-enthusiast-level $i,000 Valve Index has pumped its own respective numbers upwardly to ane,440 by 1,600 pixels per eye, or 2,880 by 3,200 pixels in total. Notwithstanding, GPUs have non kept up with the VirtualLink trend since those two were released, with many of the peak models in 2021 and 2022, released by both Nvidia and AMD, opting out of including a VirtualLink port on the dorsum of their latest cards.

If you lot decide to splurge on one of these newer headsets, you'll need a graphics carte that can keep upwards with their intense demands (80Hz refresh on the Rift S, and 144Hz refresh on the Index). This means a system that can run a graphically intensive title like Half-Life: Alyx at 144fps on two displays at once, above 1080p. Valve recommends at least a GeForce RTX 2070 Super or an AMD Radeon RX 5700 XT for the all-time experience. In short, check the specs for any headset you are considering and follow the GPU recommendations scrupulously. Substandard VR is no VR at all, only a headache and fifty-fifty nausea.

And so, Which Graphics Card Should I Purchase?

Every bit 2022 gets underway and GPUs remain scarce commodities, that reply is more convoluted than always earlier. New tech is changing the way that games interact with GPUs, and that evolution will merely continue to muddy the waters at the meridian stop while keeping both the midrange and low end in turmoil, thank you to prototype-sharpening considerations (and possible discounts).

(Photo: Zlata Ivleva)

Speaking of the top cease, right now the Nvidia GeForce RTX 3060 Ti, RTX 3070, RTX 3070 Ti, RTX 3080, RTX 3080 Ti, and RTX 3090 tin can't be beat past anything AMD has on offer on the footing of MSRP price-to-functioning alone, and then if you're looking for the most power for ability's sake, Nvidia is your go-to.

AMD, meanwhile is still holding stiff in the value-oriented midrange market, and its Radeon RX 6000 Serial launch has brought the company back into the high finish, albeit at a slightly less competitive position than Nvidia. The driver implementations will presumably get meliorate; in some games, though, they continue to plague the RX 5000 line a twelvemonth after the launch of the RX 5700 and RX 5700 XT. That may requite some buyers interruption.

(Photo: Zlata Ivleva)

Lower down the line, Nvidia and AMD barrel heads regularly, each with their own answers to sub-$1,000 PC builders that span a huge number of different bill of fare models and available options.

The GPUs we recommend at the superlative and bottom of this article span the spectrum of budget to high end, representing the full range of the all-time cards that are "available at present." Note, though: Those quote marks are very intentional.

The Elephant in the Room: Graphics Menu Pricing (and Availability) in 2022

We say "available now" because a supply-side effect that started in 2020 and rolled into 2022 has acquired limited availability of discrete graphics cards at every level of the market, from enterprise downwardly to the consumer level.

In a brief, here's what yous need to know: First upwards, it'southward no ane entity'due south "fault." Not Nvidia's, not AMD's, and certainly non Bitcoin's (on its own). The pricing situation as we know it—every bit of this writing in early on 2022—is due to a confluence of factors. And even without the pandemic, analysts in the infinite had predicted we were already on this crash course years earlier information technology actually happened.

(Photo: Zlata Ivleva)

Every industry that relies on semiconductors is feeling a supply squeeze right now, from graphics cards to the automotive industry. However, the trouble is arguably worse in GPUs. Unlike the card-shortage situation in 2017-2018, which was driven by the initial price surge in cryptocurrency, this time effectually crypto is surging at the same time that tariffs accept hit, on top of a rising in bots designed to get through Newegg and Amazon'southward shopping-cart captcha restrictions. This has lead to a situation in which cards sell out in minutes, sometimes even seconds, after they proceed auction. They're then often apace repackaged and scalped on eBay and via other venues for multiples of the list price, which certain buyers in the market have shown they are clearly willing to pay.

Many in the industry run into this particular squeeze lasting well into 2022 and across. Plus, every bit the shortages go on, the demand for discrete desktop GPUs simply compounds over time. As less people get their hands on the supply of last-generation GPUs before the side by side generation is released, the price increases continue on downward the line.

We're already catching glimpses of this, with cards like the GTX 1080 Ti going for $100 more on eBay today(Opens in a new window) than the MSRP on the twenty-four hour period information technology was launched...in 2017.

It'southward a wild world out at that place, folks.

Point is: We know that none of the prices that nosotros've discussed at an MSRP level reflects the reality of the situation for buyers today. Take this guide every bit your best objective baseline for how the carte du jour yous want should stack up on relative performance against other options. Simply as far as real-world pricing goes? Y'all could get a winning lottery ticket or a loser depending on the time of twenty-four hours you ringlet through Best Buy, Newegg, eBay, or Amazon listings. There are no guarantees in the current marketplace. When (or if) you get what you're looking for, and at what price, is anyone's best guess. The best y'all can do is use our strategies for trying to get a card at close to MSRP. Information technology's possible; information technology's just not easy.

"The All-time Graphics Cards for 2022," and then, are, to an extent this: "Whatsoever cards you tin can discover at a fair cost." And for anyone who'southward planning to wait this whole thing out for the next generation of GPUs to hitting shelves instead, expect forward to the Nvidia GeForce RTX 4090(Opens in a new window) to hopefully ease supply, when information technology launches this April 1st!