iGPU
Initialism of integrated graphics processing unit, a graphics processing unit which has been built into another major system component, such as the CPU or motherboard, and typically utilises system RAM instead of having VRAM chips of its own.
iGPU: graphics power built straight into your processor
An integrated graphics processing unit (iGPU) is a GPU etched directly onto the same silicon die as the CPU, or in some cases mounted on the same substrate without dedicated video memory. Instead of drawing from its own pool of VRAM, the iGPU allocates a portion of system RAM (typically 32 MB to 2 GB, depending on motherboard settings) as video memory. This architecture eliminates the cost and space overhead of a discrete graphics card while handling everyday display output, 2D rendering, and light 3D workloads.
Modern iGPUs appear in most mainstream CPUs: Intel's Iris Xe and Arc A-series, AMD's Radeon RDNA-based units in Ryzen processors, and Qualcomm's Adreno in mobile chips. Performance varies sharply. A current-generation desktop iGPU can manage 1440p gaming at low settings or 4K video decode, but will struggle with professional CAD rendering or sustained high-frame-rate titles. Mobile iGPUs are deliberately constrained for power efficiency. Industrial systems often rely on iGPUs specifically because they need no separate card, draw minimal power, and generate less heat in confined spaces like factory floor terminals or embedded controllers.
Common failure points and trade-offs
The main weakness of iGPU design is thermal coupling: the GPU shares the CPU's heat output and thermal solution, so sustained graphics load raises overall package temperature. If the motherboard BIOS allocates too much system RAM to video use, application performance suffers noticeably. Some industrial systems lock GPU memory allocation at firmware level to prevent configuration drift. Discrete graphics cards always outperform their integrated equivalents at the same generation, but cost, power draw, and physical size often make iGPU the only practical choice in embedded and space-constrained applications.
The term "iGPU" emerged in the 2000s as Intel and AMD began embedding GPUs into consumer processors; earlier integrated graphics were called onboard graphics or onchip graphics. In industrial settings, iGPU presence is often a purchasing criterion: CNC machines, inspection cameras, and HMI (human-machine interface) panels frequently specify systems with iGPU capability to avoid the cost and complexity of a separate video card. Firmware updates can improve driver efficiency and decode capability, making iGPU performance a moving target across product lifespans.