Hardware

GPU (graphics processing unit)

Türkçe: GPU (grafik işlemcisi)

A processor with thousands of small cores that performs the same operation on lots of data in parallel, used for graphics and AI computation.

// EXPLANATION

GPUs are many times faster than CPUs at parallel work such as matrix multiplication, which is why they are the backbone of neural-network training and inference. They have their own memory (VRAM) and talk to the CPU over PCIe.

// ALSO KNOWN AS

GPUgraphics cardgraphics processing unitvideo card

// RELATED TERMS

VRAMThe high-bandwidth memory on the graphics card itself; an AI model's weights have to fit into it. NPU (neural processing unit)A dedicated processing unit designed to accelerate neural-network computation (matrix/tensor operations) at very low energy. PCIeThe lane-based serial bus standard that connects graphics cards, SSDs and other components to the motherboard. CUDANVIDIA's software platform for programming GPUs for general-purpose parallel computing.

// ARTICLES THAT MENTION THIS TERM

ARTIFICIAL_INTEL ARTIFICIAL_INTEL November 1, 2026 Leonardo.Ai Flow State Guide: Infinite Style & Visual Lock Overcome prompt paralysis with Leonardo.Ai Flow State. We break down how latent space vector perturbation transforms a 4-word idea into an infinite scrolling waterfall of lighting, camera angles, and art styles—and how clicking "More Like This" locks your exact visual identity. 6 MIN READ #LEONARDO_FLOW_STATE

HARDWARE HARDWARE October 30, 2026 AI & Evolutionary Algorithms in Tesla Composite Metallurgy How do Tesla and SpaceX bypass decades of trial-and-error metallurgy? We explore how multi-objective evolutionary algorithms (NSGA-III), CALPHAD thermodynamics, and Quantum DFT simulations synthesize zero-thermal-expansion (ZTE) composites for 20,000 RPM Plaid motors, Giga Press castings, and Starship propulsion. 8 MIN READ #EVOLUTIONARY_ALGORITHMS

ELECTRICAL_ENERGY ELECTRICAL_ENERGY October 26, 2026 Why Global Energy Demand Is Surging: AI & Grid Bottlenecks If modern processors and appliances are more efficient than ever, why is global electricity demand surging past 30,000 TWh? We examine hyperscale AI data center rack densities, the Jevons Paradox, millisecond load transients, and the high-voltage transformer shortage. 6 MIN READ #GLOBAL_ENERGY_CRISIS

HARDWARE HARDWARE September 16, 2026 Beyond the Thermal Wall: Blackwell B200 and Liquid Cooling As an electrical and electronics engineer studying high-power semiconductor designs, you keep running into one unchanging law of nature: "Compute capacity depends not on the limits of silicon, but on the heat flux… 5 MIN READ #SMARTPHONE

SCIENCE_SOCIETY SCIENCE_SOCIETY September 14, 2026 What Is the Digital Panopticon and How Do You Escape It? Bentham's Panopticon no longer needs stone walls or a watchtower; it lives in the chips in our pockets and in data centers. Here is how we walked into the algorithmic cage, and how to get out… 5 MIN READ #HARDWARE

MOBILE MOBILE September 14, 2026 A Thermal Revolution: The iPhone 18 Pro Max Vapor Chamber As an electrical and electronics engineer examining hardware designs, the rule I pay most attention to is this: "Performance is determined not by the chip's clock speed (GHz), but by its thermal threshold." For years… 4 MIN READ #SMARTPHONE

← Whole glossary