Key Points

• GPUs evolved from simple video chips in the 1970s to powerful processors for gaming, AI, and more.
• Major companies like Nvidia, AMD, and Intel, with manufacturers like TSMC, drove innovation.
• Each generation introduced key advancements, such as 3D graphics, programmable shaders, and AI capabilities.
• Research suggests no major controversies, but competition between companies shaped development.

What Are GPUs and Why Do They Matter?

Graphics Processing Units (GPUs) are specialized chips designed to handle visual data, like images and videos, and perform complex calculations quickly. Initially created for rendering graphics in games and computers, GPUs are now used in artificial intelligence, scientific research, and even self-driving cars. Their development reflects a quest to make computers faster and visuals more realistic.

How Did GPUs Start?

In the 1970s, companies like RCA and Motorola made early video chips for arcade games and home computers, such as the Atari 2600. These chips displayed simple graphics, like blocky characters, at low resolutions.

Major Steps in GPU Development

• 1980s: Standardized graphics cards for PCs, like VGA, improved colors and resolution. ATI Technologies began making graphics hardware.
• 1990s: 3D graphics emerged with 3Dfx’s Voodoo and Nvidia’s GeForce 256, called the first GPU, which made games look more lifelike.
• 2000s: Nvidia and ATI (later AMD) introduced programmable GPUs, allowing more realistic visuals. Nvidia’s CUDA (2007) let GPUs do non-graphics tasks, like science calculations.
• 2010s: GPUs got faster and more efficient with Nvidia’s Kepler, Pascal, and Turing architectures. They added features like ray tracing for realistic lighting and became vital for AI.
• 2020s: Modern GPUs, like Nvidia’s RTX and AMD’s RDNA 2, power advanced gaming, AI, and consoles like the PlayStation 5. Intel joined with its Arc GPUs.

Who Made This Happen?

Nvidia, AMD, and Intel are the main players today, with Nvidia leading since the 1990s. TSMC manufactures the tiny, complex chips for these companies. Researchers like Jensen Huang (Nvidia’s CEO) and David Kirk (Nvidia’s scientist) were key in pushing GPU technology forward.

For a detailed history, read the full story below.

The Evolution of GPUs: A Historical Narrative

The story of the Graphics Processing Unit (GPU) is one of relentless innovation, fierce competition, and transformative technology. From humble beginnings as simple video chips to becoming the backbone of modern computing, GPUs have reshaped how we interact with digital worlds. This narrative traces their development from first principles, highlighting each generation’s major advancements, the key players, and the manufacturers who brought these chips to life.

1. The Dawn of Graphics: 1950s–1970s

In the 1950s, the seed for GPUs was planted with the need to display data visually. The MIT Whirlwind computer (Whirlwind) used a cathode-ray tube to show basic graphics, a novel concept in an era of punch cards. The first principles were simple: computers needed hardware to translate binary data into images.

By the 1970s, large-scale integration (VLSI) enabled the creation of specialized chips. Companies like RCA and Motorola led the charge:

• RCA’s “Pixie” chip (1976) produced a 62x128 resolution video signal, powering early arcade games.
• Motorola’s MC6845 (1978) laid the foundation for IBM’s Monochrome Display Adapter (MDA) and Color Graphics Adapter (CGA).
• The Atari 2600’s Television Interface Adaptor (TIA) (1977) brought graphics to home gaming.

Key Players: RCA, Motorola, Atari, and Intel.
Manufacturers: Chip designers often manufactured their own silicon, as the semiconductor industry was nascent.
Major Change: Dedicated hardware for graphics emerged, shifting rendering from software to specialized chips.

2. Standardization and Growth: 1980s

The 1980s saw graphics become a core part of personal computing. Standardized interfaces like MDA, CGA, Enhanced Graphics Adapter (EGA), and Video Graphics Array (VGA) brought higher resolutions and more colors to IBM PCs. Key developments included:

• NEC’s μPD7220 (1981), a graphics processor supporting 1024x1024 resolution.
• Hitachi’s ARTC HD63484, offering up to 4K resolution.
• Intel’s 82720 Graphics Display Controller (1983), handling 256x256 resolution with 8 colors.
• ATI Technologies, founded in 1985, released the Color Emulation Card (1986) with 16KB memory.

Silicon Graphics Inc. (SGI) also pioneered workstation graphics with its IRIS systems, setting the stage for advanced rendering.

Key Players: NEC, Hitachi, Intel, ATI, Commodore, SGI, and IBM.
Researchers: Mark Dean at IBM contributed to PC graphics standards like VGA.
Manufacturers: Texas Instruments and early foundries began supporting chip production.
Major Change: Standardized graphics cards made PCs visually richer, paving the way for broader adoption.

3. The 3D Revolution: 1990s

The 1990s were a turning point, driven by the demand for real-time 3D graphics in gaming. Companies like 3Dfx, Nvidia, and ATI transformed the industry:

• 3Dfx’s Voodoo Graphics (1996) dominated with its 3D acceleration, capturing an 85% market share by 1997. Its Glide API became a gaming standard.
• Nvidia’s NV1 (1995) and RIVA 128 (1997) set the stage for its GeForce 256 (1999), marketed as the first GPU. It integrated geometry processing and rasterization, introducing hardware transform and lighting (T&L) for faster 3D rendering.
• ATI’s 3D Rage (1995) competed as an early 3D accelerator.
• APIs like OpenGL (early 1990s) and Direct3D (1996) standardized 3D development.

Key Players: 3Dfx, Nvidia, ATI, S3 Graphics, Matrox, and PowerVR.
Researchers: Jensen Huang, Nvidia’s co-founder and CEO, drove the GeForce 256’s development.
Manufacturers: TSMC emerged as a key foundry, producing chips for GPU designers.
Major Change: The shift to 3D graphics and the introduction of the GPU concept revolutionized gaming and visual computing.

4. Programmable Shaders and GPGPU: 2000s

The 2000s saw GPUs become more flexible and powerful. Nvidia and ATI (acquired by AMD in 2006) led with programmable architectures:

• Nvidia’s GeForce 3 (2001) introduced programmable shading, enabling complex visual effects like realistic lighting.
• ATI’s Radeon 9700 (2002) supported Direct3D 9.0, enhancing graphical fidelity.
• Nvidia’s GeForce 8 series (2006) laid the groundwork for general-purpose computing.

In 2007, Nvidia’s CUDA platform (CUDA) allowed GPUs to perform non-graphics tasks, such as scientific simulations and machine learning, marking the rise of General-Purpose GPU (GPGPU) computing. OpenCL (2009), an open standard, further broadened GPU programming.

Key Players: Nvidia, AMD, Intel (with integrated GPUs).
Researchers: David Kirk, Nvidia’s Chief Scientist, was pivotal in CUDA’s development.
Manufacturers: TSMC solidified its role as the primary foundry for advanced GPUs.
Major Change: Programmable shaders and GPGPU capabilities turned GPUs into versatile computing engines.

5. AI and Advanced Architectures: 2010s

The 2010s saw GPUs evolve into powerhouses for gaming, AI, and beyond. Nvidia’s architectures advanced rapidly:

• Kepler (2012) introduced GPU Boost for dynamic performance.
• Maxwell (2014) used a 28nm process for better efficiency.
• Pascal (2016) adopted a 16nm process for the GeForce 10 series.
• Volta (2017) added tensor cores for AI and HBM2 memory.
• Turing (2018) introduced real-time ray tracing for realistic lighting.
• Ampere (2020) enhanced performance for gaming and AI.

AMD’s Radeon series progressed with Graphics Core Next (GCN) and RDNA (2019), adding ray tracing. Intel entered the discrete GPU market with its Arc series (2022).

GPUs became critical for AI, with Nvidia’s tensor cores accelerating neural network training. Applications expanded to self-driving cars, with Nvidia’s Tegra GPUs powering automotive systems.

Key Players: Nvidia, AMD, Intel, ARM.
Researchers: Jon Peddie, founder of Jon Peddie Research, documented GPU history extensively.
Manufacturers: TSMC led with advanced processes like 14nm and 7nm.
Major Change: GPUs became central to AI, machine learning, and advanced graphics like ray tracing.

6. The Modern Era: 2020s

Today, GPUs are indispensable across industries. Nvidia’s RTX series (20 series in 2018, 30 series in 2020, 40 series in 2022) leads with ray tracing and Deep Learning Super Sampling (DLSS) for enhanced visuals. AMD’s RDNA 2 (2020) powers the Radeon RX 6000 series and consoles like the PlayStation 5 and Xbox Series X/S. Intel’s Arc GPUs compete in the discrete market.

GPUs now drive gaming, AI, cryptocurrency mining, and scientific research, with applications in supercomputers and smartphones.

Key Players: Nvidia, AMD, Intel, ARM.
Researchers: Ongoing innovation continues, with figures like Jensen Huang shaping the future.
Manufacturers: TSMC’s 5nm and 3nm processes enable cutting-edge GPUs.
Major Change: GPUs are now universal computing platforms, excelling in graphics, AI, and beyond.

Key Milestones Table

Conclusion

The GPU’s journey reflects humanity’s drive to visualize and compute at ever-greater scales. From RCA’s Pixie to Nvidia’s RTX, each generation built on the last, driven by companies like Nvidia, AMD, and Intel, and enabled by manufacturers like TSMC. Researchers like Jensen Huang, David Kirk, and Jon Peddie shaped this evolution, turning GPUs into the heart of modern technology.

Key Citations:

• TechSpot: History of the Modern Graphics Processor
• Wikipedia: Graphics Processing Unit Overview
• Medium: A Brief History of GPU Development
• Jon Peddie Research: GPU History from Inception to AI