Understanding Graphics Cards

It can be difficult to decide which graphics card to purchase.


Over time, video game graphics have become more complex. These days, games are becoming increasingly photo-realistic, and with enhanced lighting, geometry, and anti-aliasing techniques, they demand a lot from your hardware.

Because of the complexity of modern graphics, a dedicated Graphical Processing Unit (GPU) is assigned the duty of rendering all graphical objects on screen, including your operating system’s graphics and your hardware accelerated apps.

Graphics cards are crucial for a variety of other purposes in addition to rendering games. For starters, these GPUs are ideal for cryptocurrency mining. They’re also required in any system that’s used for creative work, such as graphic design, video editing, or 3D modeling.

Almost all Adobe Creative Cloud programs function best with a dedicated graphics card and support GPU hardware acceleration.

This does not imply that every computer has a discrete graphics card; in fact, many low-cost or budget PCs have integrated GPUs, which are graphical processing units built into the CPU.

These GPUs are typically not very powerful, and they are unable to run many graphically demanding games. A graphics card is frequently placed on higher-end computers to ensure that the system performs better on all fronts.

Choosing which graphics card to buy can be difficult, given top competitors AMD and Nvidia introduce many devices each year. If you’re purchasing a pre-built system, such as a laptop or a branded PC, the options for which GPU to install are already reduced down to a single option, but for enthusiasts or those wishing to upgrade, the decision can be difficult.

With new games requiring DirectX 12 and Ray-Tracing compatibility, many gamers and enthusiasts are looking for a new graphics card, which is why we’re going over graphics cards today.

While the phrases graphics cards and GPUs are frequently used interchangeably, they are not interchangeable. A graphics card is the entire card including transistors, memory modules, and cooling systems, but a GPU is merely the processing unit, which is the primary silicon die.

AMD and Nvidia are the two primary companies in the GPU market. These firms create, fabricate, and manufacture GPU dies, which they normally distribute with a reference board. Third-party chip makers such as MSI, ASUS, Zotac, and Gigabyte typically produce the full graphics card. As a result, different manufacturers produce several distinct variations of the same graphics card.

While none of these chip companies will have their own GPU, they will differ in numerous elements like as cooling systems, transistor quality, and out-of-the-box overclocking. It’s preferable to go with the producer you have the greatest faith in for this portion of the process.

In terms of performance, none of these cards will have radically varied feature sets and, depending on OC, will perform similarly regardless of manufacturer. A reputable business, on the other hand, will employ high-quality VRAM chips, transistors, and well-engineered cooling systems, all of which are critical for the card’s longevity.

The base GPU, on the other hand, will be produced by AMD or Nvidia. And the model you select is the most important determinant of your graphics card’s performance. Both companies re-engineer and build new designs for these GPUs, making them not just faster but also more efficient than prior models. RTX, G-Sync, and DLSS are feature sets that have been incrementally added to GPUs.

Both organizations use a simple method for users to comprehend these criteria. On the Nvidia side, there are the popular GTX and RTX lines. GTX graphics cards are older models that do not support modern technologies like as DLSS and Ray-Tracing.

They’re also out of date now that Nvidia has switched to the RTX series. Ray-Tracing, a new lighting technology utilized in a few large games including Cyberpunk, Battlefield, and Minecraft, is supported by the RTX series of GPUs.

These are Nvidia’s most recent cards, each of which has a generation and model number linked to its name. The RTX 2060, for example, is divided into two parts: 20 (for the generation) and 60 (for the model).

The GPU design will be the same across all models in the same generation, with minor tweaks to the unit for improved performance. The model performs better when the last two values are higher. The RTX 2060 will be less powerful than the 2080, and vice versa, with the higher models’ prices reflecting this.

On the AMD side, the GPU naming method hasn’t worked out so well. However, AMD has developed a new mechanism with the latest series of graphics cards that makes things easier. Nvidia’s new RX range features a similar identifying code. The RX 5500 follows a similar pattern, with the first number 5 denoting the generation and the 500 denoting the model.

AMD’s 6000 series competes with Nvidia’s new line of GPUs, the RTX 3000 series. Both companies have GPUs that have been improved. Nvidia used the TI suffix for the GTX series until switching to Super for the RTX, while AMD used the XT suffix.

Because these are essentially slightly tweaked versions of the original model with improved performance, the easiest way to think about it is that an RTX 2060 would perform worse than an RTX 2060 Super, which would still be slower than an RTX 2070. On AMD’s side, it’s the same.

Most of these GPUs are also connected with memory modules, which means that regardless of the manufacturer, they all have the same amount of VRAM. However, while purchasing a graphics card, the amount of VRAM is also something to consider.

VRAM is similar to your CPU’s RAM in that it’s a lightning-fast storage unit that caches all of the data your GPU will need. Want to play games at higher resolutions, with more detailed textures, edit 4K video, and do 3D modeling? All of these demand a significant amount of VRAM, so consider your use case situation and determine how much you’ll need.

Nvidia and AMD both make workstation graphics cards, with Nvidia’s Quadro and AMD’s Radeon Pro. Because these workstation graphics cards are so pricey, they aren’t worth it for gaming.

These cards are also made for graphics tasks like rendering, which is why the Quadro 8000 has 48GB of GDDR6 memory, which is a VRAM capacity you’ll never reach when gaming. For recent games, an 8GB to 16GB graphics card should suffice.

There are plenty of other specifications as well, such as the amount of CUDA Cores, PCI-e Generations, and Memory Bandwidth Speed, but these are only relevant if you want to get into the details. These are especially necessary if you wish to overclock your GPU, which needs knowledge in and of itself.

All RTX 2000 and 3000 series graphics cards are available on Daraz, Aliteq, and Makura Tech online stores, so you may find one that fits your budget. On HamroBazar, you can also get older and used ones, although I wouldn’t advocate going below the RTX 2000 and RX 5000 series.

Buying a recent-generation card is always better for future-proofing, and players will miss out on many critical features that will undoubtedly define the future of gaming if they stick with the GTX series.

Source- The Kathmandu Post


The persistence of the “Learning Styles” myth


Previous Next