
AMD APU
AMD chips combining CPU and GPU on one die
Advanced Micro Devices produces these 64-bit processors that merge a central processing unit with graphics hardware. First launched in 2011, they power laptops, desktops and game consoles by sharing resources on a single silicon chip.
- Combines CPU and GPU on a single die
- First launched in June 2011
- Powers PlayStation 4 and Xbox One consoles
- Uses shared memory for CPU and graphics
- Process nodes range from 32 nm to 4 nm
What it is
The AMD Accelerated Processing Unit, previously called Fusion, is a line of microprocessors that places both a general-purpose CPU and a 3D graphics processor on the same piece of silicon. This design allows the two components to share memory and work together more closely than separate chips could. Advanced Micro Devices introduced the first models in 2011 to serve both high-performance desktops and low-power mobile devices. The approach aims to improve efficiency for tasks that benefit from parallel processing, such as video playback and light gaming, without needing a separate graphics card.
How it came about
Development began in 2006 after AMD acquired graphics maker ATI. The project faced delays due to technical challenges in merging different types of circuits at small manufacturing scales and internal debates over design priorities. After several iterations, the company unveiled the first generation in January 2011 at a consumer electronics show. Early models targeted distinct market segments, with Llano for performance and Brazos for efficiency. The brand name Fusion was later dropped following a trademark dispute, and the technology evolved through multiple generations including Trinity and Kaveri before adopting the Zen architecture.
Inside the chip
These chips use the AMD64 instruction set and integrate CPU cores with Radeon graphics hardware. Early versions relied on K10 or Bobcat cores, while later designs adopted Piledriver, Jaguar, and Steamroller architectures. The graphics units support standards like DirectX 11 and OpenCL, enabling them to handle non-graphical compute tasks. A key feature is the shared memory controller, which partitions physical RAM between the processor and graphics engine. Power management technologies adjust clock speeds dynamically to balance performance and heat output within a defined thermal limit.
How it was made
Manufacturing processes have shrunk over time, starting at 32 nanometres for early Llano and Trinity models. Later generations moved to 28 nm, 14 nm, 7 nm, and eventually 4 nm nodes. Taiwan Semiconductor Manufacturing Company produced many of these chips, particularly the low-power Kabini and Temash variants. Transistor counts grew from roughly 1.1 billion in early high-performance models to nearly 5 billion in Raven Ridge parts. Packaging formats include ball grid arrays for mobile devices and pin-based sockets like FM1 and FM2+ for desktop systems.
Where it went
Desktop computers and laptops use these processors for everyday computing, offering adequate graphics performance for casual games and media consumption. Low-power variants power netbooks, tablets, and small form-factor PCs where energy efficiency is critical. The technology also serves the gaming console market, with semi-custom versions driving the Sony PlayStation 4, Microsoft Xbox One, and their successors. This widespread adoption in consoles highlights the ability of the architecture to deliver balanced performance for fixed-hardware environments.
Versions and family
The product line includes the A-Series, E-Series, C-Series, and FX models for various performance tiers. Codenames identify specific generations, such as Llano, Trinity, Richland, Kaveri, Kabini, and Temash. Mobile and desktop versions often differ in power targets and socket types. With the introduction of the Zen microarchitecture, the branding shifted to Ryzen and Athlon with Radeon Graphics. These newer models use the G suffix for desktop parts to distinguish them from CPUs without integrated graphics, continuing the legacy of combined processing and graphics capabilities.
Why it matters
This initiative marked a significant shift in how personal computer processors were designed, pushing for tighter integration of graphics and computing logic. It helped establish heterogeneous system architecture concepts in mainstream hardware. By supplying chips for major eighth-generation game consoles, the technology reached millions of households globally. The move towards unified memory access and shared resources influenced subsequent designs in both personal computing and embedded systems, demonstrating the viability of combining different processing types on a single die.
Problems and reception
Early reviews noted strong graphics performance compared to competitors but criticised the central processing power of the first-generation Llano chips. Users and analysts appreciated the improvements in CPU speed with the Trinity release. Some criticism arose regarding the short lifespan of certain sockets, such as FM1, which limited upgrade paths for early adopters. Despite mixed reactions to raw CPU speeds, the integrated graphics capabilities were consistently praised for providing better value than separate entry-level graphics cards.




Questions people ask about AMD APU
What does an AMD APU do?
It combines a central processing unit and a graphics processing unit on one chip. This allows the device to handle both general computing tasks and visual rendering without needing a separate graphics card, saving space and power.
When was the first APU released?
AMD announced the first generation in January 2011 and launched the initial units on June 14 of that year. These early models included Llano for high-performance needs and Brazos for low-power devices.
Which game consoles use AMD APUs?
Semi-custom versions of these chips power the Sony PlayStation 4 and Microsoft Xbox One. They are also used in the subsequent PlayStation 5 and Xbox Series X|S consoles, providing the core processing and graphics capabilities for these systems.
How has the manufacturing process changed?
Early models like Llano were built on a 32 nm process. Later generations moved to smaller nodes, including 28 nm, 14 nm, 7 nm, and 4 nm. This shrinkage allowed for more transistors and better power efficiency over time.
Good reading and sources
- Checking your browserAnandTech, Article
- The AMD A8-3500M APU Review: Llano Is UnleashedTom's Hardware, Article
The facts on this page come from the Wikipedia article "AMD APU", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryChip rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.