Graphics Jargon Buster: A Graphical Glossary
Looking to refresh your businesses’ computers? Want to brush up on your sales skills when providing GPUs to customers? There’s a lot of lingo when it comes to graphics cards, and with new features, technologies, and software being introduced all the time, it’s more important than ever that you’re clued in.
Read on to discover the most common and relevant terms you’ll want to understand.
Boost Speed/Clock: The figures displayed with these terms reflect the max speeds (measured in MHz) the graphics card can boost or overclock to, usually when needed during particularly intense applications or moments of gameplay.
Clock Speed: The clock speed indicates the speed of the graphics card, measured in MHz. Generally speaking, the higher the clock speed, the better the output of the card, and the better performance users can expect in games and applications.
CUDA Cores: Exclusive to NVIDIA, CUDA (once known as Compute Unified Device Architecture) is a parallel computing platform that enables applications outside of games to efficiently use the graphics card’s power. Learn more via NVIDIA’s blog.
DLSS (Deep Learning Super Sampling): DLSS is a recent AI rendering technology that makes use of NVIDIA’s Tensor Cores (see below for an explanation on Tensor Cores) and a deep learning neural network to enhance gaming fidelity and performance. It does this by using AI rendering to produce an image that’s similar in quality to native resolutions, but rendering a mere fraction of the native image.
Learn more about DLSS in our dedicated article on DLSS and FSR.
FSR (FidelityFX Super Resolution): Similar to NVIDIA’s DLSS (see above for an explainer on DLSS), AMD’s FSR upscales imagery using cutting-edge technologies, boosting framerates. Unlike DLSS, FSR is software-based, and is usable on a range of GPUs, including recent NVIDIA products.
GDDR: GDDR stands for Graphics Double Data Rate, and is the specialised memory type used in graphics cards. The current GDDR version is GDDR6, providing 16 Gbps data rates – twice that of GDDR5.
GPU: The shorthand term for Graphics Card, GPU stands for Graphics Processing Unit.
Infinity Cache: An AMD exclusive technology, Infinity Cache is a new cache level, enabling high-bandwidth performance. The cache is seen by the entire graphics core, enabling rapid data access and efficient processing of graphical data.
Memory Bandwidth: Memory Bandwidth measures the rate of data reads and writes by the GPU, and is typically measured in GB/s (Gigabytes per second).
Memory/VRAM Size: The amount of memory, known as VRAM (Video Random Access Memory), the graphics card possesses. VRAM processes any image requests by the system, and acts as a buffer between the CPU (Central Processing Unit) and the display outputting the images rendered by the GPU.
Memory Speed: The speed of the memory (VRAM) on the GPU, typically measured in MHz.
Microsoft DirectX® 12: DirectX is a group of application programming interfaces (APIs) that offer developers low-level access to the components of a GPU. The benefit of this is easier access to graphical effects and features for developers, and richer visuals for gamers.
PCIe® Gen 4 (4.0): You’ll hear a lot about PCIe® 4.0 these days, and for good reason. PCIe® slots are the motherboard’s spaces for PCIe®-compatible products, most commonly graphics cards or high-speed storage devices.
PCIe® 4.0 is the latest standard of PCIe technology, doubling the speeds over the previous generation.
PCIe® 5.0 and 6.0 are already in the process of being created, and both subsequently double the bandwidth and transfer speeds of the preceding generation.
Ray Tracing: A form of graphical rendering that accurately simulates light, realistically reflecting and dispersing light throughout a scene in a game or visual media. Historically, game developers have used static or baked-in lighting that typically doesn’t respond to changes in the environment (day/night cycles are an example), in order to save on performance costs.
With AMD and NVIDIA’s latest graphics cards, this performance cost is minimised, as hardware technologies included in these contemporary products focus on providing real-time reflections.
RT Cores: NVIDIA’s term for the cores found on their RTX GPUs, RT cores are designed to specifically accelerate computations involved in ray tracing, eliminating the typical performance costs associated with such intensive mathematical computations.
RTX: The contemporary nomenclature for NVIDIA’s ray tracing-enabled graphics cards, RTX stands for Ray Tracing Texel eXtreme.
Smart Access Memory: An AMD technology, Smart Access Memory enables an AMD CPU to access the full amount of VRAM offered by the GPU via the bandwidth of PCIe® lanes, removing bottlenecks and increasing performance.
TDP: Thermal Design Power (or Point depending on who you ask), commonly known as TDP, measures the maximum level of heat output by the GPU. This is commonly measured in Watts (W).
It’s important to note that TDP is not the GPU’s average power draw; the GPU may draw this level of power when used intensively, but likely won’t maintain such metrics consistently unless used in very specific use-cases.
TDP is a useful metric for deciding what power supply (PSU) is needed to support the GPU’s maximum output.
Vulkan: Vulkan, similar to DirectX, is a low-level API used by developers to access the lower functionalities of GPUs in order to improve game efficiency and performance, as well as deliver visual effects.
Looking for your next GPU refresh? Browse ASUS Graphics now.
Read on to discover the most common and relevant terms you’ll want to understand.
Boost Speed/Clock: The figures displayed with these terms reflect the max speeds (measured in MHz) the graphics card can boost or overclock to, usually when needed during particularly intense applications or moments of gameplay.
Clock Speed: The clock speed indicates the speed of the graphics card, measured in MHz. Generally speaking, the higher the clock speed, the better the output of the card, and the better performance users can expect in games and applications.
CUDA Cores: Exclusive to NVIDIA, CUDA (once known as Compute Unified Device Architecture) is a parallel computing platform that enables applications outside of games to efficiently use the graphics card’s power. Learn more via NVIDIA’s blog.
DLSS (Deep Learning Super Sampling): DLSS is a recent AI rendering technology that makes use of NVIDIA’s Tensor Cores (see below for an explanation on Tensor Cores) and a deep learning neural network to enhance gaming fidelity and performance. It does this by using AI rendering to produce an image that’s similar in quality to native resolutions, but rendering a mere fraction of the native image.
Learn more about DLSS in our dedicated article on DLSS and FSR.
FSR (FidelityFX Super Resolution): Similar to NVIDIA’s DLSS (see above for an explainer on DLSS), AMD’s FSR upscales imagery using cutting-edge technologies, boosting framerates. Unlike DLSS, FSR is software-based, and is usable on a range of GPUs, including recent NVIDIA products.
GDDR: GDDR stands for Graphics Double Data Rate, and is the specialised memory type used in graphics cards. The current GDDR version is GDDR6, providing 16 Gbps data rates – twice that of GDDR5.
GPU: The shorthand term for Graphics Card, GPU stands for Graphics Processing Unit.
Infinity Cache: An AMD exclusive technology, Infinity Cache is a new cache level, enabling high-bandwidth performance. The cache is seen by the entire graphics core, enabling rapid data access and efficient processing of graphical data.
Memory Bandwidth: Memory Bandwidth measures the rate of data reads and writes by the GPU, and is typically measured in GB/s (Gigabytes per second).
Memory/VRAM Size: The amount of memory, known as VRAM (Video Random Access Memory), the graphics card possesses. VRAM processes any image requests by the system, and acts as a buffer between the CPU (Central Processing Unit) and the display outputting the images rendered by the GPU.
Memory Speed: The speed of the memory (VRAM) on the GPU, typically measured in MHz.
Microsoft DirectX® 12: DirectX is a group of application programming interfaces (APIs) that offer developers low-level access to the components of a GPU. The benefit of this is easier access to graphical effects and features for developers, and richer visuals for gamers.
PCIe® Gen 4 (4.0): You’ll hear a lot about PCIe® 4.0 these days, and for good reason. PCIe® slots are the motherboard’s spaces for PCIe®-compatible products, most commonly graphics cards or high-speed storage devices.
PCIe® 4.0 is the latest standard of PCIe technology, doubling the speeds over the previous generation.
PCIe® 5.0 and 6.0 are already in the process of being created, and both subsequently double the bandwidth and transfer speeds of the preceding generation.
Ray Tracing: A form of graphical rendering that accurately simulates light, realistically reflecting and dispersing light throughout a scene in a game or visual media. Historically, game developers have used static or baked-in lighting that typically doesn’t respond to changes in the environment (day/night cycles are an example), in order to save on performance costs.
With AMD and NVIDIA’s latest graphics cards, this performance cost is minimised, as hardware technologies included in these contemporary products focus on providing real-time reflections.
RT Cores: NVIDIA’s term for the cores found on their RTX GPUs, RT cores are designed to specifically accelerate computations involved in ray tracing, eliminating the typical performance costs associated with such intensive mathematical computations.
RTX: The contemporary nomenclature for NVIDIA’s ray tracing-enabled graphics cards, RTX stands for Ray Tracing Texel eXtreme.
Smart Access Memory: An AMD technology, Smart Access Memory enables an AMD CPU to access the full amount of VRAM offered by the GPU via the bandwidth of PCIe® lanes, removing bottlenecks and increasing performance.
TDP: Thermal Design Power (or Point depending on who you ask), commonly known as TDP, measures the maximum level of heat output by the GPU. This is commonly measured in Watts (W).
It’s important to note that TDP is not the GPU’s average power draw; the GPU may draw this level of power when used intensively, but likely won’t maintain such metrics consistently unless used in very specific use-cases.
TDP is a useful metric for deciding what power supply (PSU) is needed to support the GPU’s maximum output.
Vulkan: Vulkan, similar to DirectX, is a low-level API used by developers to access the lower functionalities of GPUs in order to improve game efficiency and performance, as well as deliver visual effects.
Looking for your next GPU refresh? Browse ASUS Graphics now.
