29 Mar, 2025

Video cards

The Graphics Card (GPU)

The graphics card, or GPU (Graphics Processing Unit), is an essential component in a gaming PC. It is responsible for rendering graphics, processing images, and accelerating 3D tasks, directly influencing the performance and visual quality of games. In this article, we will explore in detail the role of graphics cards in gaming and the important considerations to take into account when choosing the right GPU for you. We will focus exclusively on cutting-edge gaming video cards.

Factors to Consider When Choosing a Graphics Card:

ARCHITECTURE

Graphics cards are based on various architectures and naming conventions, including NVIDIA GeForce RTX, AMD Radeon RX, and Intel ARC. Each architecture has its own features and proprietary technologies that impact performance and functionality. In this guide, we won’t cover Intel ARC GPUs as we do not consider them competitive with NVIDIA and AMD cards yet.

Recently, NVIDIA launched the new 5000 series, based on the Blackwell architecture. These new GPUs promise significant improvements in ray tracing and AI, offering exceptional performance in graphically intensive games at high resolutions. AMD’s new 9000 series GPUs, based on RDNA 4 architecture, have significantly improved ray tracing capabilities and computational power, creating tougher competition with NVIDIA in the gaming space.

CORE CLOCK AND MEMORY CLOCK

When choosing a GPU, one of the most important specifications to consider is the Core Clock, followed by the Memory Clock. These two parameters directly impact performance and game smoothness.

The Core Clock refers to the speed at which the graphics processor operates. The higher the core clock, the faster the GPU can process the necessary graphical computations to render games and other visually demanding applications. A higher core clock translates to better performance, allowing for higher frame rates and smoother gameplay.

On the other hand, the Memory Clock indicates the speed at which the GPU’s video memory (VRAM) operates. A higher memory clock means that the GPU can access the data stored in VRAM faster, improving overall rendering speed and reducing the risk of stuttering or lag during gameplay.

In conclusion, both the core clock and memory clock are important for ensuring optimal performance in a graphics card. “Founder” editions offer a solid base, while custom GPUs can offer additional benefits through aggressive overclocking, enhanced cooling, and other complementary features. The choice depends on each user’s specific needs and personal preferences. These two values can be overclocked using specific software, and the extent to which they can be “boosted” depends on the manufacturer/edition of the GPU.

POWER LIMIT

The power limit of a GPU is an important feature that determines the maximum amount of electrical power the GPU can consume during operation. Limiting power consumption helps keep GPU temperatures within safe limits, reducing the risk of overheating and extending its lifespan.

This limit is usually set by the manufacturer for several reasons:

Thermal Control: Limiting power consumption helps keep the GPU temperature within certain bounds, reducing the risk of overheating. This is essential to ensure that the GPU runs reliably and extends its lifespan.
System Power Compatibility: The power limit ensures that the GPU does not exceed the power limits of the system, preventing compatibility issues and ensuring the GPU runs properly even in systems with limited power.
Overclocking: Some advanced users may want to adjust the power limit to allow for more aggressive overclocking of the GPU, increasing performance at the expense of higher power consumption. However, it's important to be careful not to exceed safe limits to avoid damaging the GPU.

CUSTOM GPU AND FOUNDER EDITIONS

Custom GPUs are those made by external partners like ASUS, MSI, Gigabyte, Zotac, and others. These partners take the reference GPU design from NVIDIA or AMD and customize it by adding custom heat sinks, modified PCBs, advanced cooling solutions, and sometimes even factory overclocking or dual BIOS. Custom GPUs can offer slightly better performance than reference versions and may come with additional features such as RGB lighting and advanced management software.

On the other hand, Founder or reference editions are those made directly by NVIDIA or AMD. These cards are based on the reference GPU design and are typically the first to launch when a new GPU generation is released. However, they usually feature factory clocks that are almost identical to Custom versions.

The main difference between custom GPUs and Founder editions is customization. Custom GPUs offer more personalization options and can be optimized for specific performance and features. Founder editions, on the other hand, provide a reliable, well-built solution directly from NVIDIA or AMD, ideal for those who want a plug-and-play solution without much customization.

In summary, Founder editions provide a solid base, while custom GPUs can offer additional benefits through aggressive overclocking, better cooling, and other extra features. The choice depends on the user's specific needs.

Video Memory (VRAM) The amount and speed of video memory are crucial for handling texture and graphic data. The more VRAM, the better the performance at higher resolutions and more advanced graphic settings. Nowadays, even low-end GPUs come with at least 8 GB of VRAM, which is sufficient for running games at medium-high graphics settings in high resolutions.

Advanced Graphics Technologies and Features Here are some advanced technologies that an NVIDIA RTX GPU of the latest generation may have:

Real-time Ray Tracing: Ray tracing is an advanced rendering technique that simulates the realistic behavior of light, creating more accurate and natural shadows, reflections, and refractions, significantly enhancing the visual experience.
DLSS (Deep Learning Super Sampling): DLSS is a technology that uses artificial intelligence to improve performance and graphics quality in games. By using deep learning, DLSS can reduce the resolution of games without sacrificing much visual quality, thus allowing for better performance.
Variable Rate Shading (VRS): Variable Rate Shading is a technology that enables the graphics card to allocate more or fewer rendering resources to different parts of the image depending on their visual importance. This allows for optimized GPU resource usage, improving performance without compromising overall visual quality.
Resizable BAR: Resize Bar is a technology introduced by NVIDIA that allows the CPU to directly access a larger portion of the GPU’s video memory, improving performance and reducing latency in games and applications. This allows for greater efficiency in using GPU resources.
DirectX 12 Ultimate and Vulkan: NVIDIA GPUs support the latest graphics APIs, DirectX 12 Ultimate and Vulkan, which offer a wide range of advanced features, including real-time ray tracing, variable rate shading, and other technologies to ensure maximum compatibility and performance in the latest games.

On the AMD side, the latest AMD RADEON RX GPUs also come with the following advanced features:

AMD FidelityFX Super Resolution 3.0 (FSR): This is an AI-based upscaling technology developed by AMD (equivalent to NVIDIA's DLSS). FSR 3.0 uses advanced machine learning algorithms to enhance resolution and image quality in games. Along with FSR 3.0, AMD also implements the "AMD Frame Generation" (or Fluid Motion Frames) technology, which focuses on creating intermediate frames through frame interpolation, improving motion fluidity in games and reducing stuttering.
AMD Smart Access Memory (SAM): SAM is a technology that allows the CPU to directly access the GPU’s video memory, improving overall system performance. This increases available bandwidth and reduces latency, contributing to better gaming performance.
Variable Rate Shading (VRS)
DirectX 12 Ultimate and Vulkan

NVIDIA vs. AMD GPUs Choosing between NVIDIA and AMD often comes down to personal preference and specific needs. Generally, NVIDIA GPUs are known for superior ray tracing, AI-driven technologies like DLSS, and performance in high-end gaming scenarios, especially for games that rely heavily on ray tracing.

On the other hand, AMD cards usually offer better value for money, especially in the mid-range and budget segments. They perform very well at standard gaming resolutions like 1080p and 1440p and tend to be more efficient in terms of power consumption.

For those who want the best performance in ray tracing and AI, NVIDIA is the preferred choice. However, for a better price-to-performance ratio, especially in games without heavy ray tracing demands, AMD might be the better option.

GPU Drivers The drivers for GPUs play a crucial role in ensuring optimal performance and stability. NVIDIA drivers are often considered more stable and reliable, offering a better overall experience, especially in high-demand gaming scenarios. AMD is improving, but NVIDIA is generally seen as superior in terms of driver quality and software integration.

Streaming For streamers, NVIDIA GPUs offer excellent compatibility with streaming software like OBS Studio and Streamlabs OBS, along with dedicated hardware for video encoding (NVENC), which provides superior performance and video quality compared to AMD’s VCE.

Bottleneck The term “bottleneck” refers to a component in a PC that limits the overall performance of the system. A CPU-GPU bottleneck occurs when one of the two components (CPU or GPU) is running significantly slower than the other, hindering the overall performance of the system.