Everything else on a gaming computer is secondary to the graphics card. Without a powerful GPU to push pixels, even the most powerful CPU cannot accomplish much. Even though no single graphics card is ideal for everyone, we will provide options for every budget and mentality below. We have you covered whether you’re looking for the fastest graphics card, the best value, or the best card at a specific price.
In contrast to our GPU benchmarks hierarchy, which ranks cards based solely on performance, our list of the best graphics cards considers the entire package. Price, availability, performance, features, and efficacy are all essential, but their relative importance is subjective.
The good news is that the long night of GPU shortages and exorbitant prices is about to end. The profitability of cryptocurrency mining plummeted, and graphics card prices fell an additional 6% on average last month. All major GPUs are now in stock at online retailers, with the majority available at or below their official MSRPs.
How We Test the Best Graphics Cards
The best method for determining the performance of a graphics card is to eliminate all other bottlenecks, or at least as much as possible. Core i9-12900K CPU, MSI Z690 DDR4 motherboard, 32GB Corsair DDR4-3600 CL16 memory, Crucial P5 Plus 2TB SSD, Cooler Master PSU, case, and CPU cooler.
We conduct tests at the three most popular gaming resolutions, 1080p, 1440p, and 4K, with medium and ‘ultra’ settings. We utilize reference cards whenever possible for all of these tests, such as Nvidia’s Founders Edition models and AMD’s reference designs. However, most midrange and low-end GPUs lack reference models; in some cases, we only have factory-overclocked cards for testing. In such instances, we do our best to select cards that closely match the reference specifications.
Each graphics card undergoes the same testing protocol. After launching the game, we run one pass of each benchmark to “warm up” the GPU, followed by at least two passes at each setting/resolution combination. If the difference between the two runs is less than 0.5%, we use the faster of the two runs. If the difference is significant, we repeat the test at least twice more to determine what “normal” performance should be.
The RTX 3070 Ti, RTX 3070, and RTX 3060 Ti are all expected to perform within a narrow range. The RTX 3070 Ti is approximately 5% faster than the RTX 3070, which is about 5% faster than the RTX 3060 Ti. If we observe games with clear outliers (i.e., performance is more excellent than 10% higher for the cards just mentioned), we will retest the cards in question to determine the “correct” result.
Due to the time required to test each GPU, updating drivers and game patches will inevitably impact performance. We periodically retest a handful of sample cards to ensure our results are still accurate, and if not, we retest the affected game(s) and GPU (s). In the coming year, we may also add games to our test suite if one is released that is popular and conducive to testing — see our benchmark for what makes a good game for our selection criteria.
10 Best Graphic Cards For PC Gaming
Our list consists almost exclusively of cards from the current generation, except the GTX 1660 Super. This is nearly as fast as the RTX 3050, and while it lacks ray tracing hardware, Amazon’s recent price of $210 makes up for this.
This is why the RTX 3090 Ti is positioned above the RX 6950 XT, and the RTX 3080 is positioned above the RX 6800 XT in the preceding table. Our subjective rankings below also consider price, performance, and features. Others may have a slightly different perspective, but every card on this list merits your consideration.
1. Best Overall: GeForce RTX 3080 ULTRA GAMING
Nvidia’s GeForce RTX 3080 with Nvidia’s current Ampere architecture remains our top recommendation. This now applies to all RTX 3080 cards, including the RTX 3080 Ti and RTX 3080 12GB — based on current prices. Currently, the best deals are $740 for a 10GB card, $760 for a 12GB model, and $800 for a 3080 Ti.
Obtain this card if you’re serious about maximizing all graphics settings and playing at 4K or 1440p. It’s more than you’ll typically need for 1080p gaming unless you’re playing the most recent ray-tracing games, in which case DLSS support should improve performance. DLSS also functions without ray tracing and is currently available in over 200 games, making it a feature that is definitely worth a premium.
AMD provides the universal FSR 2.0 as an alternative to DLSS, but it is not present in as many games. Intel’s XeSS, on the other hand, resembles DLSS, but only for Intel GPUs, which we are still awaiting outside the entry-level Arc A380.
The looming arrival of Nvidia’s next-generation RTX 40-series GPUs is the primary concern with purchasing an RTX 3080-series card. We’ve seen several leaks and rumored specifications, and we anticipate that Nvidia’s CEO will provide additional details during his GTC keynote on September 27. If past performance indicates, the most powerful RTX 40-series GPUs may launch a week or two later. Thus, spending $700 or more on a GPU whose generation will soon be obsolete is questionable.
- Real boost clock: 1755 MHz; Memory detail: 10240 MB GDDR6X.
- Real-time ray tracing in games for cutting-edge, hyper-realistic graphics.
- Triple HDB fans offer higher performance cooling and much quieter acoustic noise.
- All-metal backplate & adjustable ARGB
- Excellent performance
- Good bang for the buck
- Can do 4K ultra at 60 fps or more
- Prices remain above MSRP
- Requires 320W or more power
- Overkill for 1080p displays
2. Best AMD GPU: Radeon RX 6800 XT CORE Gaming Graphics
The best cards for Team Red are Navi 21 GPUs from AMD, such as the Radeon RX 6800 XT. The RX 6800 XT provides a massive increase in performance and features compared to the previous generation RX 5700 XT, as well as ray tracing support (via DirectX Raytracing or VulkanRT) thanks to its RDNA 2 architecture. The RX Radeon 6900 XT and RX 6800 are also worthy of consideration, as their prices have dropped significantly over the past few months. The 6900 XT, for instance, now costs 17% more for roughly 5% greater performance, while the RX 6800 loses 7% performance and costs 7% less.
Before its release, the enthusiast community affectionately dubbed the Navi 21 GPU “Big Navi,” and we received precisely what we desired. It is more than twice as large as the previous Navi 10, with twice as many shader cores and twice as much RAM. The RX 6800 XT has a 300W TBP, slightly lower than the RTX 3080’s 320W TBP. Clock speeds are also increased to the range of 2.1-2.4 GHz (depending on the card model), and AMD accomplished all of this without substantially increasing power requirements: the RX 6800 XT has a 300W TBP.
The enormous 128MB Infinity Cache contributes significantly to AMD’s performance. AMD claims that this increases the adequate bandwidth by 119%. Few games, if any, require more than 16GB, so the 6800 XT excels in this regard.
What can be disliked? Due to AMD’s lack of BVH traversal hardware, ray tracing performance is subpar (it uses GPU shaders for that). Tensor cores and DLSS are also absent, although FSR 2.0 partially compensates. As with Nvidia’s top GPUs, the upcoming RDNA 3 presents the most significant cause for concern. Waiting for the RX 7000-series to arrive (before the end of the year) is preferable to spending $600 or more on a GPU from two years ago if you do not already own one of AMD’s top GPUs.
- Memory Speed:16 GBPS
- The Speedster series exemplifies a modern aerodynamic style though clean and elegant design. It is a thoughtful design with the sole purpose of maximizing airflow to improve cooling and performance.
- A revolutionary new memory architecture that redefines how to deliver higher levels of performance and efficiency for 4K gaming. AMD RDNA 2 architecture elevates and unifies the gaming
- AMD RDNA 2 architecture elevates and unifies the gaming experience from performance to visuals across consoles and PC.
- New Ray Accelerators, handling the intersection of rays, deliver high performance hardware accelerated ray tracing
- RDNA2 architecture provides excellent performance
- Easily handles 4K and 1440p
- Lots of VRAM for the future
- FSR 2.0 needs wider adoption
- Weaker ray tracing performance
- Still overpriced
3. Excellent Performance: XFX Speedster QICK319 AMD Ultra Gaming Graphics Card
AMD’s Navi 22 and RX 6700 XT are derived from the Navi 21 GPU by paring down the various functional units to create a smaller die that can be sold at lower prices. The RX 6750 XT is identical to the RX 6750, with a slight increase in clock speeds, memory speeds, and power consumption, making it about 5% faster overall but 12% more expensive.
The 6700 XT has the same GPU cores as the RX 5700 XT from the previous generation. Still, significantly faster clock speeds and more cache give it a roughly 25% performance boost (at higher settings and resolutions, at least). During testing, the RX 6700 XT achieved clock speeds over 2.5GHz while playing games with the reference card’s factory settings. Models with factory-overclocked clock speeds can push this closer to 2.7GHz without overheating the GPU.
The RX 6700 XT traded blows with the RTX 3070 and RTX 6060 Ti in our performance tests. It is slightly faster than the latter and slower than the former, but its average price of $420 is below both. Nonetheless, if we include most games with DLSS or ray tracing, the 6700 XT falls behind the 3060 Ti and resembles a 3060 competitor.
This card has risen in our overall rankings due to its competitive online pricing. It is currently available at prices beginning just below the suggested retail price. Keep an eye on the more recent RX 6750 XT, as it may prove to be a superior option if prices continue to fall. We also do not anticipate a replacement for the RX 7000 until 2023, although we could be wrong.
- AMD RDNA2 Architecture & Hardware Ray Tracing
- Triple Double Ball Bearing Fans
- Ghost Thermal Design
- Extreme Stability & Overclocking
- Great 1080p and 1440p performance
- Plenty of VRAM
- Excelling price to performance ratio
- Weaker RT performance
- FSR can't defeat DLSS
- Incoming RDNA2 refresh
4. Good Mainstream Performance: PowerColor Hellhound AMD Radeon RX 6650 XT Graphics Card
The Navi 23 architecture represents AMD’s response to the RTX 3060. Usually, we’d anticipate a 32 CU variant of Navi 22, dubbed the RX 6700 non-XT. Still, AMD reduced CU counts, memory interface width, and Infinity Cache sizes to create a more minor, less expensive chip with comparable performance. (Note that the Radeon RX 6700 now exists, with 10GB of VRAM and a significantly higher price.)
Even though the memory bus has been cut in half to a mere 128 bits, performance is slightly higher than the previous-generation RX 5700 XT, which is impressive considering the memory bus has been reduced by half. Despite the understandable concern regarding the 8GB of VRAM, the RTX 3060 may be the superior option in certain circumstances. Nonetheless, it is surprising how much even a 32MB Infinity Cache boosts performance when examining memory bandwidth. This is essentially a more minor chip than Navi 10, built on the same TSMC N7 node, and it provides 10–15% better 1080p framerates.
However, there are instances in which it struggles; ray tracing is significant. Several games with DXR (DirectX Raytracing) support could not even achieve 20 frames per second at 1080p. The RTX 3060 from Nvidia was roughly twice as fast without DLSS and an additional 40% faster with DLSS Quality mode. FSR doesn’t address this issue, as it provides the same performance boost to AMD, Nvidia, and Intel GPUs. The RX 6650 XT and RX 6600 XT feel like a letdown compared to the other Big Navi chips, which boasted impressive amounts of VRAM.
This is reflected in current online pricing, which has helped somewhat improve the situation. The RX 6600 XT’s MSRP is $379, and the RX 6650 XT’s MSRP is $399, but both are available for as little as $300. AMD is not scheduled to replace its current midrange offerings until 2023.
- Video Memory: 8GB GDDR6
- Stream Processor: 2048
- Game Clock: 2486 MHz (OC) / 2410 MHz (Silent)
- Boost Clock: 2689 MHz (OC) / 2635 MHz (Silent)
- Memory Clock: 17.5 Gbps
- Minimum System Power: 600W
- Faster than 3060 and RX 5700 XT
- Power efficient design
- Good 1080p performance
- Available below MSRP
- Only 8GB VRAM on a 128-bit bus
- Poor ray tracing performance
- Expensive for 1080p
5. Good for 1440p Gaming: NVIDIA GeForce RTX 3060 Graphics Card
After testing the GeForce RTX 3060 Ti, we concluded that it might be the best of Nvidia’s Ampere GPUs. It has the same features as the other GPUs in the 30-series and starts at just $399. In theory, it sold out as quickly as every other new graphics card. However, things have improved, and the lowest price we can find right now is approximately $405, which is still higher than MSRP but getting closer.
In our testing, the 3060 Ti outperformed the previous-generation 2080 Super in every game we ran. Additionally, it was only 9 percent slower than the RTX 3070 while costing 20 percent less. If you are still using an older GTX 1070 or RX Vega 56, the 3060 Ti is up to twice as fast and, in some cases, even faster in the most recent games.
The only legitimate concern is the absence of VRAM. Currently, 8GB is presently sufficient for most games, but some titles are beginning to exceed that threshold. Of course, you can reduce the texture quality and may not even notice the difference, but you’ll regret it in your heart. (Not really; high settings frequently appear identical to ultra settings.)
The RX 6650 XT and RX 6600 XT from AMD provide stiff competition for the 3060 Ti. Nvidia’s component is still faster, especially in ray-tracing games, but the RX 6650 XT is currently $150 less expensive than the 3060 Ti.
- GeForce RTX 30 Series GPUs deliver the ultimate performance for gamers and creators. They’re powered by Ampere—NVIDIA’s 2nd gen RTX architecture— with new RT Cores, Tensor Cores, and streaming multiprocessors for the most realistic ray-traced graphics and cutting-edge AI features.
- Experience today’s biggest blockbusters like never before with the visual fidelity of real-time ray tracing and the ultimate performance of AI-powered DLSS. RTX. It’s On.
- NVIDIA DLSS is groundbreaking AI rendering that boosts frame rates with uncompromised image quality using the dedicated AI processing Tensor Cores on GeForce RTX. This gives you the performance headroom to crank up settings and resolutions for an incredible visual experience. The AI revolution has come to gaming.
- NVIDIA Reflex delivers the ultimate competitive advantage. The lowest latency. The best responsiveness. Powered by GeForce RTX 30 Series GPUs and NVIDIA G-SYNC monitors. Acquire targets faster, react quicker, and increase aim precision through a revolutionary suite of technologies to measure and optimize system latency for competitive games.
- Take your creative projects to a new level with GeForce RTX 30 Series GPUs. Delivering AI-acceleration in top creative apps. Backed by the NVIDIA Studio platform of dedicated drivers and exclusive tools. And built to perform in record time. Whether rendering complex 3D scenes, editing 8K video, or livestreaming with the best encoding and image quality, GeForce RTX GPUs give you the performance to create your best.
- Beats the 2080 Super for $300 less
- Good overall value (fps/$)
- Great for RT at 1440p with DLSS
- Still overpriced at present
- 4K is a a stretch even with DLSS
- 8GB might not be enough VRAM long term
6. Best Overall Value: Radeon RX 6600 CORE Gaming Graphics Card
The Radeon RX 6600 reduces the 6600 XT’s strengths by a small margin. It is approximately 15% slower than the RTX 3060 overall, but in our testing, it was still 30% faster than the RTX 3050. It’s also priced to sell, with the cheapest cards beginning at just $250, such as this XFX RX 6600 card at Amazon
That is significantly less than AMD’s official MSRP of $329, which seemed a bit high at launch — not that we saw those prices in significant quantities until recent months. But with cards shipping well below MSRP, this is the best value on the market overall.
The RX 6600 competes with the RTX 3050 and previous generation RTX 20-series and GTX 16-series GPUs in the competitive midrange graphics card market. Our testing achieved near-RTX 2070 performance, at least in non-ray-tracing scenarios. With ray tracing enabled, it performed poorly, barely averaging 30 frames per second in our DXR test suite at 1080p medium and falling 20% behind Nvidia’s RTX 2060.
If you are not concerned with ray tracing, the RX 6600 is worth looking at. AMD’s Infinity Cache does wonders for an otherwise somewhat underpowered GPU, and the card only requires 130W, significantly less than competing GPUs.
- Memory Speed:14 GBPS
- Chipset: AMD RX 6600
- Memory: 8GB GDDR6 AMD RDNA 2 architecture elevates and unifies the gaming
- Boost Clock: Up To 2491MHz
- Cooling: XFX Speedster SWFT210 Dual Fan Cooling
- Power efficient
- Runs 1080p max settings and 60fps
- Typically costs less than MSRP
- Not good for ray tracing
- Can't match the RTX 3060
- Only 8GB VRAM
7. Best mid-range graphics cards: GeForce RTX 3070 Graphics Card
The Nvidia GeForce RTX 3070 is the best graphics card for most users due to its outstanding performance. It is comparable to the RTX 2080 Ti without being prohibitively expensive, bringing 4K gaming to the mass market for the first time. Before this card, 4K gaming was out of reach for many people.
The RTX 3070 accomplishes this without requiring compromises in settings for most games. Yes, this reasonably-priced graphics card powering your PC allows you to play your favorite AAA games in 4K with maximum settings without experiencing any performance degradation. The RTX 3070 is at the top of our list: it combines a low price with 4K gaming on high-quality settings. Additionally, we appreciate that it does not have the same massive cooler as the RTX 3080, allowing it to have a smaller form factor.
- Chipset: NVIDIA GeForce RTX 3070
- Video Memory: 8GB GDDR6
- Memory Interface: 256-bit
- Output: DisplayPort x 3 (v1.4a) / HDMI 2.1 x 1
- Digital maximum resolution – 7680 x 4320
- Amazing performance
- Best value graphics card today
- Awesome ray tracing performance
- Same inflated prices as Turing
- Required 12-pin power connector
8. AMD's Fastest GPU: XFX Speedster MERC319 RX 6950XT Black Gaming Graphics Card
The RX 6950 XT represents the pinnacle of performance for the RDNA2 architecture, outperforming the older RX 6900 XT by an average of 9%. AMD set the MSRP at a relatively high $1,099 at launch, but cards are already selling for significantly less. The RX 6950 and 6900 XT are essentially the same GPU, but the RX 6950 has a faster 18Gbps GDDR6, a higher power limit, and slightly faster GPU clocks.
The RX 6950 XT has slightly more GPU cores than the RX 6800 XT, and when combined with the difference in clock speeds, it is approximately 15% faster overall. However, the 6950 XT is roughly 50 percent more expensive than the 6900 XT, and the 6900 XT is about $200 less expensive, so choosing AMD’s penultimate GPU over-the-top card is not a terrible idea.
The RX 6950 XT is the fastest GPU available for 1080p and 1440p gaming but falls short at 4K. The usual caveats regarding lower ray tracing performance and the lack of DLSS support also apply, and while FSR 2.0 appears promising, it is not widely supported by games yet. Nvidia continues to offer the best DXR/RT experience, though you don’t need a ray tracing to enjoy games.
The 6950 XT may appeal to those only interested in the fastest AMD GPU. Compared to the existing RX 6900 XT, it is not revolutionary, but we did not expect it to be. The upcoming RDNA 3 GPU launch, likely to occur before the end of 2022, is of more significant concern. AMD promises a 50% improvement in performance per watt, but we’ll have to wait and see what this means for actual gaming. Rather than upgrading to the 6950 XT in the RDNA 2 life cycle, it makes much more sense to wait a few months to see how things pan out.
- Chipset: AMD RX 6950 XT
- Memory: 16GB GDDR6
- Cooling: XFX MERC 319 Triple Fan
- Boost Clock: Up To 2368MHz
- Excellent overall performance
- Lots of VRAM and Infinity Cache
- Fastest in non-RT workloads
- Good SPECviewperf results
- High starting MSRP
- Slower than Nvidia in RT performance
- RDNA 3 coming later this year
9. Good 1080p Graphics With DXR and DLSS: MSI Gaming GeForce RTX 3050 Graphics Card
Nvidia attempted to create a “budget” RTX 30-series graphics card with the GeForce RTX 3050, its $250 suggested retail price places it squarely in the mainstream category. It is also selling for $300 or more, which is better than the launch prices but not as low as we would like, given that it was 7% slower than the previous generation RTX 2060 in our tests.
In general, we’d rather pay for an RTX card than a GeForce GTX 1660 Super or RX 5500 XT 8GB, though the former is currently back in stock and selling for closer to $200. (see below for our next pick).
In our testing, the RTX 3050 was approximately 15% faster than a GTX 1660 Super. Additionally, it can legitimately run ray-tracing games and supports DLSS. That is more than we can say for AMD’s RX 6500 XT, which likely should have sacrificed RT support for more VRAM and bandwidth. In contrast, AMD’s RX 6600 (above) was 30% faster in standard games and only 13% slower in DXR games while costing $25 less than the least expensive 3050 we can find at this time.
The biggest issue is that $300 or more is still a lot to pay for mainstream performance, and we hope to see prices drop to MSRP within the next few months. Those who don’t care about ray tracing and DLSS have a better option with AMD, so this is only for those who prefer to stick with Nvidia, even if the available performance isn’t all that impressive.
- Chipset: GeForce RTX 3050
- Video Memory: 8GB GDDR6
- Memory Interface: 128-bit
- Output: DisplayPort x 3 (v1.4a) / HDMI 2.1 x 1
- Digital maximum resolution: 7680 x 4320
- Cheapest RTX GPU so far
- Full RTX and DLSS feature set
- Still provides 8GB VRAM
- Slower than the RTX 2060
- Also loses to RX 6600
- Online prices still inflated
10. Budget Gaming for Under $200: Radeon RX 6500 XT Gaming Graphics Card
Radeon RX 6500 XT utilizes AMD’s RDNA 2 GPU, notably newer than the budget GPUs that are typically used. With only a 64-bit memory interface, 16MB Infinity Cache, an x4 PCIe link, no video encoding support, and only two display outputs, the Navi 24 chips were severely downsized. Numerous potentially exciting features have been removed.
Nonetheless, if cost is your primary concern, the RX 6500 XT starts at $176 on Amazon, making it less expensive than most other options. For less than $200, your only other options are the GTX 1650 Super, GTX 1650, or Intel Arc A380.
The RX 6500 XT is our default recommendation for sub-$200 budgets. We’d much rather have a GTX 1660-series card, RTX 2060, or even the previous generation’s RX 5500 XT 8GB. Still, these cards cost more than $200 — unless you shop on eBay, but buying a used graphics card is risky, as many miners are likely selling cards that have been heavily utilized over the past two years.
- Memory Speed:18.0 Gbps.Digital Max Resolution:7680×4320
- Country Of Origin : China
- Stream Processor: 1024
- Game Clock: 2650Mhz
- Boost Clock: 2820Mhz
- Handles 1080p medium
- Actually affordable
- Current generation architecture
- 4GB VRAM is limiting
- Only two video ports
- x4 PCIe link
- Still needs a 6-pin power connector
- Terrible RT performance
- No video encoding hardware
How to Choose the Best Graphics Card for Your Gaming PC
When purchasing, building, or upgrading a gaming PC, processor speed is one of the most critical factors. It would help if you had a CPU that could handle the demanding resource requirements of modern games.
However, the graphics card is at least as necessary. The graphics card is responsible for rendering the gaming environment on the monitor. Without a powerful graphics card for PC gaming, your favorite game may stutter, freeze, or fail to load.
But with dozens of graphics card models available, which one should you choose? Let’s attempt to deconstruct that question.
The Brands Making Graphics Cards
AMD and NVIDIA are the two most prominent manufacturers of graphics cards for PC gaming. Intel intends to release its own graphics card, but it is unknown when this will occur.
AMD and NVIDIA have different naming conventions for their graphics card lines, making comparisons challenging (but not impossible).
AMD currently sells gaming graphics cards under the brand name Radeon, with three primary product lines: RX 5000 Series, RX Vega Series, and RX 500 Series. Regarding model names, R stands for Radeon, and X indicates a higher-end model than, for example, the Radeon 500 series. The first number is the card’s generation number, and the remainder is AMD’s ranking of the card compared to others in the same line. For instance, the RX 5700XT provides superior performance to the RX 5500 XT. The XT indicates higher performance than a card lacking that designation.
AMD Radeon RX 5700 XT is one of the company’s most popular graphics cards. According to a review on the technology website Techradar.com, it is “absolutely an excellent graphics card. It stands out even in the most competitive graphics card market ever.”
The names of the graphics cards manufactured by NVIDIA can also be confusing. Currently, the company markets its graphics cards under GeForce and has two naming schemes: GTX and RTX.
GTX is NVIDIA’s standard line, whereas RTX cards include dedicated hardware to run the company’s RTX suite of gaming features that enhance lighting, reflections, and shadows in supported games, a sharper image, and increased FPS.
Following either GTX or RTX will be four digits. The first two indicate the card’s generation, while the second two indicate its relative position compared to other cards of the same type. The RTX 2080, for example, offers better performance than the RTX 2060.
Some cards also have a “TI” or “Super” designation following the number, denoting an improvement over the non-TI or non-Super version (but not as good as a card with a higher number). For instance, a card with the designation 1070 TI is superior to a 1070 but inferior to a 1080.
When it comes to graphics cards, NVIDIA currently reigns supreme. The NVIDIA GeForce RTX 2080 TI is the best premium graphics card on the market in 2020, according to the website T3.com. However, this performance comes at a cost. Expect to pay at least $1,400 for a 2080 TI. Prices vary, however.
Graphics cards for PC gaming can be distinguished based on their model numbers and rankings. Still, to determine which card best suits your needs, it’s essential to be familiar with the card’s specifications. Here’s a summary:
- Graphics card memory amount: One of the most important specifications, especially when playing a game at a high resolution or with maximum settings, is the amount of graphics card memory. If you are playing a game at 4K resolution, you will need a card with at least 8 GB memory.
- Ports: Ensure that the card you are considering is compatible with your monitor. If not, you will need to buy an adapter or even a new monitor. Typically, newer monitors have HDMI or DisplayPort connections, whereas some older models only have DVI.
- Clock speed: The clock speed, measured in megahertz (MHz), indicates the speed of a graphics card’s cores. The cores render graphics, so the faster the processing, the higher the GPU’s clock speed. Numerous games opt for overclocking or running the card faster than the manufacturer-specified speed. Clock speed is essential, but it is not the only factor affecting a card’s performance.
- CUDA Cores / Stream Processors: CUDA Cores are parallel processors comparable to a CPU’s dual- or quad-core configuration. The cores are in charge of processing the data entering and leaving the card. For a meaningful comparison, you must compare the number of cores within the same architecture instead of architectures of different types. Comparing core counts across architectures or brands of cards is not a reliable indicator of performance differences.
- TFLOPS / GFLOPS: TFLOPS, or trillions of floating-point operations per second, measures a GPU’s theoretical maximum performance. Alternatively, it may be expressed as GFLOPS or billions of FLOPS. TFLOPS is computed by multiplying the number of cores by the clock speed and then by two. TFLOPS can indicate how much faster one chip is compared to another within the same architecture.
- Memory speed/bandwidth: Faster memory can result in a faster card.
- RT / Tensor Cores: Ray tracing-focused RT cores and machine-learning-focused Tensor Cores are relatively new screen-detailing technologies typically found on high-end NVIDIA cards. They have potential but limited game support at present.
But Will it Fit?
There are few things more disappointing than purchasing the graphics card of your dreams only to discover that it does not fit your case. It’s an essential but frequently overlooked detail.
When upgrading a graphics card for PC gaming, consider the dimensions of the card’s length, height, and thickness. Typically, graphics cards are available in half-height (slim), single-slot, dual-slot, and even triple-slot configurations. Most gaming cards manufactured today are full-height and require two expansion slots. In addition, a card with a large heatsink and fan can obstruct access to an adjacent slot.
In addition, most gaming cards are powered by one or more 6- or 8-pin PCIe power connectors that come in various configurations. You will need to upgrade if your power supply lacks the proper connectors. Using an adapter to connect to SATA or Molex connectors is not recommended.
TDP, or Thermal Design Power, is an additional game card specification that should be considered. TDP is a wattage-based indicator of the card’s power consumption during normal operation. When adding a new graphics card, you may need to upgrade your power supply depending on the number. In general, a 600W power supply is sufficient for all but the most powerful graphics cards, and 800W is enough for nearly any card, even when overclocked.
For 1080p gaming, acquire a card with at least 6GB, preferably 8GB or more. If you play with all the settings turned up or install high-resolution texture packs, you will require more memory. And if gaming at extremely high resolutions, such as 4K, you should have more than 8GB.
If you have a tighter budget and can’t spend more than $150, this XFX Radeon RX 6400 graphics card is a good option. The RX 6400 is a bit slower than the GTX 1660 Super and the RX 6500 XT, but it is still capable of running most games on a 1080P monitor with at least medium settings.
If you are an avid gamer, a video editor, or if you’ve owned your GPU for more than four years, you should upgrade your GPU first. In certain situations, it is preferable to upgrade the CPU first because it is more cost-effective, has a longer lifespan, and controls every aspect of the system besides the graphics card.
A GPU with 2GB is more than adequate for general use, but gamers and creative professionals should aim for at least 4GB of GPU RAM. The amount of memory required for a graphics card ultimately depends on the desired game resolution and the games themselves.
GeForce RTX™ 30 Series, powered by Ampere, NVIDIA’s 2nd generation RTX architecture, delivers the ultimate performance, ray-traced graphics, and AI-powered DLSS for gamers and creators.
This article reviews the most important considerations when selecting a graphics card for a gaming PC. These tips should assist you in purchasing the optimal graphics card.
If you wish to increase your understanding of graphics cards, GPUs, and video cards, it is worthwhile to explore our other resources.
We are committed to assisting users in locating the ideal graphics card for their budget and specific requirements.
- All Post
- Computer Components
Any portable hard drive, SSD, or flash drive will allow you to store, transport, and back up your files. Obtaining…
Everything else on a gaming computer is secondary to the graphics card. Without a powerful GPU to push pixels, even…
End of Content.