How to choose the perfect graphics card for a powerful gaming experience?
If you're looking to buy a graphics card but don't know where to start, it can be nearly impossible to choose the right one. In this video, you'll learn everything you need to know about graphics cards. To satisfy all tastes, I'll cover both the basics and more advanced details. So, without wasting any more time, let's get started.
First, let me explain the terminology you need to know to avoid confusion. It may sound a bit complicated, but don't worry, you'll understand everything by the end of this video. What are DLSS, FSR, and XESS? These three technologies found in modern graphics cards do roughly the same thing. They upscale low-resolution images to high-resolution images using artificial intelligence or algorithms.
For example, you may not be able to run a game smoothly at 4K resolution because your graphics card isn't powerful enough. These three technologies upscale the resolution to 4K by processing the game's graphics at 1080p and then addressing any glitches using artificial intelligence. Depending on your settings, this may cause slight image distortion, but it improves gaming performance.
The leader in this performance-enhancing technology is NVIDIA. NVIDIA's DLSS technology has been more successful than its competitors in terms of quality and performance thanks to artificial intelligence. DLSS can only be used on NVIDIA cards. On the other hand, FSR is a performance-enhancing technology developed by AMD. While DLSS can only be used on NVIDIA cards, as I just mentioned, things are a bit more complicated with AMD's FSR technology, and it varies between versions. For example, while FSR 3.
Version 0 also supports Nvidia graphics cards, while FSR 4.0 only supports RX9000 series graphics cards. XESS is a performance upscaling technology developed by Intel, a combination of DLSS and FSR. In addition to Intel graphics cards, XESS also supports many Nvidia and AMD graphics cards, but it performs best on Intel graphics cards.
I think you understand upscaling techniques. Now let's talk about frame generation. What is frame generation? Frame generation, just like upscaling techniques, is used to improve game performance. However, it works differently. While upscaling techniques use artificial intelligence to increase resolution, frame generation analyzes frames in the game to add new frames using artificial intelligence.
This way, games become smoother. Its impact on performance is much greater than that of resolution upscaling techniques. In fact, performance increases so dramatically that NVIDIA's multi-frame generation technology can sometimes increase the frame rate by three to four times. Of course, while doing so, it can also cause artifacts.
Moreover, the worst part is that it significantly increases input lag. This makes it impractical for competitive gaming. AMD's frame generation technology is called AFMF, and Intel's is called XESS. These technologies may not be as effective as Nvidia's multi-frame generation technology, but they can cause less input lag.
Now, let's talk about ray tracing. What is ray tracing? I'd say it's the reason DLSS and the frame-generating techniques I just mentioned exist. Ray tracing produces more realistic images and reflections by calculating the interaction of light with objects in a scene. As you might imagine, since this technology requires a lot of calculations, graphics card power alone isn't enough.
Therefore, we need to leverage techniques like image upscaling and frame generation to get good performance from ray tracing. Ray tracing has been used for years in the animation and film industries. In gaming, NVIDIA launched a stellar ray tracing campaign with its RTX cards in 2018.
Nvidia cards offer better ray tracing performance than their competitors, but AMD is improving ray tracing performance with each new generation. Intel is also trying to compete, but overall, its mid-range graphics cards suffer from poor ray tracing performance and compatibility.
It's a type of core found in a graphics card. Developed by NVIDIA, these cores enable the graphics processing unit to deliver high performance in tasks that require high processing power, such as graphics processing and artificial intelligence. The faster and more numerous these cores are, the higher the performance. CUDA is available on NVIDIA graphics processing units.
AMD's CUDA cores are equivalent to the Stream cores, and Intel's XE cores are equivalent. Of these three cores, Nvidia's graphics performance generally outperforms that of the other two companies. As you can see, Nvidia has a significant advantage over other brands in terms of graphics card technology.
Okay, now you know everything you need to know about graphics card technology. Let me tell you how to choose the right graphics card. Which company is right for you? Nvidia, AMD, and Intel. These three companies are the most popular GPU manufacturers. There are many differences between the cards from these companies.
When purchasing a graphics card, it's very important to choose the right manufacturer. Let me explain. NVIDIA is the most popular graphics card manufacturer. It's known for its technologies like ray tracing, DLSS, frame rate generation, and CUDA. NVIDIA drivers are typically more stable than those from other manufacturers.
Therefore, they are considered more reliable. Their power consumption is also excellent. Of course, for all their advantages, they are expensive. Furthermore, there has been a significant shortage of stock recently with the launch of the RTX 5000 series. This is actually one of the reasons for their higher price compared to the manufacturer's suggested retail price in many countries.
Meanwhile, AMD is another competitor to Nvidia. It has alternatives to DLSS and frame-transforming technologies, but they don't match Nvidia's success. Their pricing varies from country to country, but they are able to compete with Nvidia thanks to their generally reasonable prices. In terms of raw gaming performance, some AMD cards offer significantly better performance than similarly priced Nvidia cards.
I'd like to emphasize that this is a competition only in gaming performance. NVIDIA far outperforms AMD in computing-intensive areas such as ray tracing, content creation, streaming, and artificial intelligence thanks to CUDA technology. Additionally, there's a significant bias against AMD's graphics cards due to their high power consumption and past driver issues, such as the Whatman bug.
These problems are virtually nonexistent today, so it's safe to say that these biases are meaningless. Besides these two problems, there's a new player in the market: Intel. Intel manufactures entry-level and mid-range graphics cards that focus on price-to-performance ratios. Intel also has image enhancement technologies that are alternatives to DLSS, and they offer excellent performance.
Its power consumption is also reasonable, although there may be some firmware issues due to its recent release, it performs very efficiently. So much so, that we see significant increases in GPU performance in just one month thanks to driver updates from motherboard manufacturers. So, while GPUs and graphics cards are manufactured by Nvidia, AMD, or Intel, the graphics cards you buy in stores are usually produced by third-party brands like ASUS, MSI, and...
Sapphire. These brands import GPUs from these three manufacturers and then sell them after making modifications in areas such as cooling, clock speeds, and design. In fact, there are a variety of cards, including entry-level, mid-range, and high-end, even within the same brand. This means, for example, that not all ASUS cards are the same.
You know the brands. Now, let's move on to one of the most confusing, yet also one of the most important, topics: graphics card naming systems. Nvidia, AMD, and Intel all have different types of GPUs.
To meet the needs of more users, the company divides its cards into entry-level, mid-range, and high-performance cards. This is where graphics card naming comes in. You'll need to understand this naming to choose the card that best suits your budget and usage. Let's start with Nvidia. What does GTX or RTX mean on the left? Actually, the term GTX or RTX refers to the technology base.
Currently, cards that support advanced technologies like ray tracing and DLSS are part of the RTX series. What do the first two numbers on the left mean? These numbers indicate which series the card belongs to. There are series such as the RTX 20, 30, 40, and even 50 series. The higher the number, the more recent the technology used in the card.
For example, you'll get lower power consumption, higher performance, and additional software features. In fact, technologies like multi-frame generation are exclusive to the RTX50 series. What do the last two numbers mean? The last two numbers indicate the performance the card can deliver.
Things are clearer here. 50, 60, 70, 80, 90. The higher the number, the higher the category. This means that the RTX 4070 is more powerful than the 4060. Within the same series, performance and price increase with higher numbers. But beware: a newer card doesn't necessarily mean a more powerful one.
For example, the RTX 3070 is slightly more powerful than the RTX 4060 Ti or GTX 1080 Ti, and the Old Legend performs similarly to the RTX 3060. This means that it's not enough to just look at the naming scheme; you should also review practical performance tests. We'll discuss this in more detail later. What about super or TI? The terms TI or super are typically used at the end of GPU model names for cards whose performance is between two other models.
For example, there is a gap between the RTX 4070 and RTX 4080. To bridge this gap, cards like the RTX 4070Ti are being released that are more powerful than the 4070, but cheaper than the 4080. The name "Super" may appear in some card names. These Super cards are enhanced and more powerful versions of the standard versions of cards in the same series.
For example, there's about a 20% performance difference between the RTX 4070 and the RTX 4070 Super. In short, the Ti and Super suffixes were added to provide more options and performance ratings within the same series. Next up is AMD. Frankly, AMD's way of naming cards is very confusing, especially in the older series.
But don't worry, they've moved to a more organized naming system, similar to that of NVIDIA cards, such as the 70-series AMD RX9000 series XT. You might think it's the same thing NVIDIA explained, just with the numbers changed. Now, let's get to the more complicated part. How does the naming system work on AMD cards? RX.
This term has been in use for a long time. It's the generic name for AMD graphics cards. The first number indicates the card's series, such as the RX 5000, 6000, or 7000 series. The second number indicates the card's class. For example, RX6700 refers to the 700th class of the 6000 series.
If the last two digits are 50, this card is a different version of the regular card. For example, you might think the RX6750 is a higher-performance version of the RX6700, but overall, there aren't significant performance differences between these two cards. Their performance is similar. Typically, these cards aren't worth the price difference. XT and XTX.
You might think these cards are the Ti cards in Nvidia's graphics cards, meaning they are the more powerful version of the same card. For example, the RX6700 XT is more powerful than the RX6700. Now let's move on to Intel cards. Intel is a newcomer to the graphics card market, so their names are a bit strange. For example, the most powerful graphics card in the previous generation of Intel cards is called the ARK A770, while the most powerful graphics card in the current generation is called the ARK B580.
As you can see, there's little correlation between the naming of their generations. Therefore, it's not possible to be certain about Intel's card naming scheme yet. Its user base isn't large, but it's important to note that Intel is improving itself very quickly.
Now that you know what the terms on graphics card names mean, let's look at other important things you need to know when choosing a graphics card. Resolution. As you know, resolution refers to the number of pixels on the display. The most common resolutions today are 1080p, 1440p, and 2160p, or 4K. As resolution increases, the load on the graphics card increases, resulting in lower frame rates during gameplay and increased system resource consumption.
So, if you want to play games at higher resolutions, you need to choose a powerful graphics processing unit that can handle this load. Video RAM (VRAM) is the memory of the graphics card itself. Don't confuse it with random access memory (RAM), which is system memory. This memory is where the graphics processing unit temporarily stores data. The more VRAM, the more data the graphics card can process.
However, if your VRAM is less than required, you may experience lower FPS, stuttering, and significant performance issues. The likelihood of experiencing these issues is high given the high VRAM usage in modern games, such as The Last of Us. Of course, optimizing the game you plan to play is also crucial.
Games with poor optimization can cause problems with video memory (VRAM) usage. We often encounter such problems with games ported from consoles to PCs. By the way, as resolution increases, so does VRAM usage. Therefore, I recommend choosing a graphics card with at least 8GB of VRAM if you plan to play at 1080p, and at least 12GB if you plan to play at 1440p. That's tremendous power.
Raw power is the actual performance delivered by a graphics processing unit (GPU) without auxiliary technologies like DLSS. Therefore, the performance you get from a game without any additional software reflects the raw power of your graphics card. This power is especially important for professional use cases such as rendering, modeling, and computationally demanding tasks.
Video memory (VRAM) and raw power are complementary factors. Bottlenecks occur when one is strong and the other is weak. For example, the RTX 3070 Ti has more raw power than the 4060 Ti, but we can't see its full performance because it only has 8GB of VRAM. On the other hand, the 4060 Ti has 16GB of VRAM, which is very good, but its power can sometimes be insufficient.
One card that strikes this balance perfectly is the RX7700 XT. It offers a great balance with a generous amount of raw power and 12GB of VRAM. I highly recommend this card. Power consumption: The amount of power a graphics card draws from the power supply is very important. As power consumption increases, the GPU temperature rises.
So, you need to choose graphics cards with better cooling. To find this value, we usually look at the graphics cards. Total Graphics Power (TGP) refers to the maximum power the GPU can draw. Total Board Power (TBP) refers to the amount of power a graphics card draws when all of its components are included, including the fan, lighting, video memory (VRAM), and so on.
What about TDP, or Thermal Design Power? The TDP value is mostly related to the cooling element. This is more commonly used than the other two, but TGP and TBP give the clearest results. You can also visit the Tech PowerUp website to see the recommended power supply rating for your graphics card. However, keep in mind that the rating provided by this site may sometimes be close to the maximum, and other times it may be exaggerated.
I think the best option is to ask others on the forums for help. PCI Express version. PCI Express is the interface through which the motherboard exchanges data with the graphics card. This means that despite the processing power of the graphics card, you may experience slow performance if the graphics card cannot transfer data to the motherboard quickly enough.
But is this really something to always worry about? No, it only becomes important in certain special cases. Terms like PCI Express X4, X8, and X16 refer to the actual data bus width. On the other hand, terms like PCI Express 3.0, 4.0, and 5.0 show how fast the data bus can transfer data.
However, some graphics cards come with reduced bus bandwidth. For example, the RX6500 XT only uses the PCI Express x4 lane. If you use this card with an older motherboard that only supports, say, PCI Express 3.0, the bandwidth will be significantly reduced, and you'll notice a significant drop in frame rates for cards like the RX6500 XT. PCI Express 4.0 support is essential.
What about high-performance cards? Cards like the RTX 4090 and RX7900 XTX typically use the X16 lane. For these cards, the performance difference between PCI Express 3.0 and 4.0 is minimal, as the bandwidth is sufficient. In short, if you're purchasing a GPU with a low data bus, such as the RX6400 or 6500 XT, it's crucial that your motherboard supports PCI Express 4.0.
In the mid-range and high-end segments, PCI Express version doesn't make much of a difference, but a newer motherboard is always better. Additionally, since most modern graphics cards are PCI Express 4.0 x16 compatible, you just need to make sure your computer's components are compatible. Now you know what to look for when choosing a graphics card, but how do you know how well it actually performs? The best way to find out is to watch performance tests on YouTube.
You can watch graphics card comparison videos. This makes it easier to choose the right graphics card if you're torn between two. You can also view detailed reviews of each card individually. These videos don't just focus on gaming performance testing.
Tests are also available for production loads, such as video editing and 3D rendering. This way, you can learn about general-purpose hardware performance, not just gaming performance. You can also get feedback from experienced users by creating threads in the hardware forums. You'll usually get quick and helpful responses.