What was the first graphics card and when was it released? See the details here

The first video card released for the PC was the Monochrome Display Adapter (MDA), created by IBM in 1981 along with the launch of the IBM PC 5150.

Capable of displaying only monochrome text at 80 × 25 character resolution with no graphics support, the MDA ushered in the era of discrete video adapters—cards that plug into ISA expansion slots, not chips integrated into the motherboard.

Almost simultaneously, IBM also launched the Color Graphics Adapter (CGA), the first card to bring true color and graphics to the PC world. This distinction between embedded and dedicated graphics became the basis of the modular architecture that defines personal computers to this day.

Embedded vs. Dedicated: two approaches

It is essential to understand the distinction between integrated graphics chips and dedicated cards. The onboard video chip is soldered directly to the motherboard and shares RAM and processor memory with the system, significantly reducing performance in intensive graphics tasks. The offboard board has its own processor (GPU) and dedicated memory, working independently and freeing up computer resources.

At the beginning of the PC/XT era, all components were integrated into the motherboard. With the AT standard, expansion cards emerged, allowing the user to choose external components of higher quality and performance. The open architecture of the IBM PC thus made it possible for third parties to manufacture boards with much greater capabilities.

The first steps: from IBM to Hercules

O Monochrome Display Adapter (MDA) it was the first PC video adapter capable of displaying only monochrome text at 80 × 25 character resolution, without graphics support. Just one year later, Hercules Computer Technology revolutionized the market with the Hercules Graphics Card (HGC), which combined the MDA text standard with a high-resolution (720 × 348 pixels) bitmap graphics mode. HGC had 64 KB of video memory—16 times more than MDA—making it a widely supported de facto standard, especially for dual-monitor setups used by programmers.

IBM launched the Color Graphics Adapter (CGA) in 1981 as its first color display capable board, equipped with 16 KB of memory and support for two main modes of text and graphics. Although CGA was notorious for limitations, it actually offered up to 16 colors at resolutions up to 640×200 pixels, with text and graphics modes that could be displayed on CRT RGB monitors or NTSC-compatible televisions. In practice, visual restrictions varied depending on the mode, balancing colors and resolution to ensure performance.

The breakthrough came in 1984 with the Enhanced Graphics Adapter (EGA).

The SuperEGA HiRes allowed the user to create up to 512 custom characters, supported 80×60 text modes and was equipped with a proprietary CMOS chip (V151). Another highlight was the support for multiple monitors and simultaneous use of frequencies, as well as compatibility with standard IBM software and professional applications: AutoCAD, Lotus, WordPerfect, WordStar, among others. The package included drivers for a wide range of programs, considerably expanding the potential of the equipped PCs.

These “Super EGA” models confirm how the EGA standard not only brought 16 selectable colors from a palette of 64 and 640×350 pixels, but inspired a movement of innovation among alternative manufacturers, who sought to deliver more resolution, more compatibility and advanced functions to the productivity and graphic design niche of the second half of the 1980s.

Finally, in 1987, the Video Graphics Array (VGA) revolutionized the market with support for resolutions of 640×480 pixels, up to 256 simultaneous colors, single-chip architecture and the VGA connector standard, which lasted decades. This consolidated the PC as a multimedia platform.

The first professional graphics accelerator

In 1984, IBM introduced the Professional Graphics Controller (PGC), considered the first PC graphics system with dedicated processing. The PGC consisted of a set of three cards, including an Intel 8088 processor at 8 MHz, 320 KB of dedicated RAM and support for 640×480 pixels with 256 colors at 60 Hz — far superior to CGA and EGA

Focused on CAD and professional environments, it freed the CPU from heavy graphics computing, something revolutionary for the time. The high cost restricted its use to specialized niches, but it served as a precursor to modern dedicated GPUs. It was discontinued in 1987, with the advancement of VGA and the 8514 standard.

The era of 3D acceleration

In the 1990s, 2D and 3D acceleration became popular with cards such as the S3 ViRGE, ATI Rage and Matrox Mystique, as well as the pioneering 3dfx Voodoo, which focused exclusively on 3D graphics.

But the real change came in 1999, when NVIDIA released the GeForce 256, titled the “world’s first GPU”. This single chip integrated morphing and lighting (T&L) functions directly onto the board, decoupling graphics tasks from the CPU and supporting Direct3D 7. This revolution enabled the advanced graphics and parallel processing that define today’s gaming and computing.

The first modern GPU

NVIDIA popularized the term “GPU” in 1999 when it launched the GeForce 256 on October 11, marketed as “the world’s first GPU.” The company defined GPU as “a single-chip processor with integrated transformation, lighting, triangle configuration/cropping, and rendering engines capable of processing at least 10 million polygons per second.”

The GeForce 256 included features such as hardware T&L (Transform and Lighting) in the graphics processor instead of offloading these tasks to the CPU. With a GPU manufactured by TSMC using a 220nm CMOS process, the card offered a notable leap in 3D PC gaming performance and was the first 3D accelerator fully compatible with Direct3D 7.

Collector impresses with more than 1000 video cards, 2500 CPUs, 400 motherboards and 300 PCs

Over the years, GPUs have become increasingly fundamental to the evolution of the technology market in general. Currently, we see this with everything involving the AI ​​market, with supercomputers responsible for maintaining services that have become essential to the daily lives of millions of people around the world, such as ChatGPT.