AAA vs Hombre: The lost next-generation Amiga that could have changed Commodore forever

In the early 90s the Amiga stood at one of the most important crossroads in its history. Just a few years earlier it had been the most advanced multimedia computer available to ordinary users. When the Amiga 1000 appeared in 1985 it stunned the industry with capabilities that PCs could barely approach. Smooth hardware scrolling, rich digital sound, advanced graphics and a multitasking operating system made the machine feel like something from the future. Artists, video editors, game developers and hobbyists quickly adopted it, and the Amiga built a reputation as one of the most creative computer platforms of its era. But technology was evolving quickly. By the beginning of the 90s IBM-compatible PCs were rapidly catching up. Faster Intel processors, increasingly capable graphics cards and the growing popularity of Microsoft Windows were shifting the balance of power in the personal computer market. Meanwhile the world of graphics was beginning to change as well. New companies such as Silicon Graphics were demonstrating powerful 3D workstations, while the gaming industry was moving toward real-time polygon graphics. Commodore’s engineers knew that the Amiga could not remain unchanged forever. If the platform wanted to stay relevant, it needed a new generation of hardware.

Inside Commodore two very different ideas about the future began to form. One group of engineers believed the Amiga should evolve gradually, preserving its unique architecture while improving performance. Their project became known as AAA, the Advanced Amiga Architecture. Another group believed the time had come for a complete redesign that would abandon many aspects of the original system in favor of newer technologies. Their concept eventually became known as Hombre. These two projects represented two very different visions for the future of the Amiga. One focused on continuity and compatibility, while the other aimed for a bold leap into the next generation of graphics computing. To understand why these ideas were so different, it helps to remember how the original Amiga was designed. Unlike most computers of the 80s, the Amiga did not rely solely on its central processor. Instead it used a set of custom chips that worked alongside the CPU to handle specific tasks. The three most important were Agnus, Denise and Paula. Agnus controlled memory and managed the system’s DMA channels, Denise handled graphics output and display generation, and Paula managed audio and input/output functions. Together these chips allowed the Amiga to perform complex multimedia tasks without overloading the main processor. The result was a machine that could display smooth animations, play digital sound and manipulate graphics far more efficiently than many competitors.

This custom chipset architecture became the heart of the Amiga’s identity. Over the years Commodore improved it through several revisions. The original OCS chipset was followed by ECS, which added higher resolutions and other refinements. Later came AGA, introduced in the Amiga 1200 and Amiga 4000, which expanded the color palette to over sixteen million colors and allowed up to 256 colors on screen simultaneously. For many users AGA felt like a major step forward, but engineers inside Commodore already knew that it would not be enough for the long term. One of the main limitations of the Amiga architecture was its use of planar graphics. In this system pixel data is stored across several separate bitplanes rather than being stored together as a single value. This approach worked extremely well for the types of 2D graphics common in the 80s. It allowed efficient animation, palette manipulation and other visual tricks that game developers loved. However, planar graphics became less practical when developers started experimenting with 3D rendering. Techniques like texture mapping and per-pixel shading were far easier to implement with chunky pixel formats, where each pixel is stored as a complete value in memory. PCs and emerging graphics hardware were beginning to move in that direction, leaving the Amiga’s design looking increasingly dated.

Memory bandwidth was another growing problem. The Amiga’s custom chips and CPU all shared the same chip RAM, which meant they competed for access to memory. As resolutions and color depths increased, this shared memory system could become a bottleneck. Finally there was the question of CPU performance. Motorola’s 680×0 processors were elegant and powerful, but by the early 90s they were becoming expensive and slower compared with the rapidly advancing world of RISC processors. The AAA project was Commodore’s attempt to modernize the existing architecture without throwing away what made the Amiga special. Development began around the end of the 80s, when engineers started designing a new generation of custom chips that could replace the aging Agnus and Denise components. AAA aimed to deliver higher resolutions, deeper color, faster memory access and significantly improved graphics performance while maintaining compatibility with the huge library of existing Amiga software.

The planned AAA chipset included several new chips with names such as Andrea, Linda, Mary and Monica. Andrea would manage memory control and DMA operations, Linda would act as the primary graphics processor, Mary would handle system integration and bus management, and Monica would serve as an advanced blitter responsible for fast graphics operations. Together these components would create a far more powerful multimedia system than previous Amiga chipsets. AAA promised several major improvements. Display resolutions could potentially reach levels such as 1024×768 or even 1280×1024, which would have been impressive for consumer hardware of the early 90s. The chipset would support full 24-bit color, allowing true-color images rather than palette-based displays. The blitter would become significantly faster and more capable, enabling more complex graphical operations and improved animation performance. Most importantly, AAA was designed around a wider 64-bit memory bus that could dramatically increase bandwidth compared with earlier systems.

Despite these impressive features, AAA still followed the traditional Amiga philosophy. The system retained planar graphics and remained tied to Motorola processors. For many engineers and developers this continuity was a major advantage. Existing software could continue to run, and programmers would not need to abandon the techniques they had spent years mastering. From a practical standpoint AAA offered a relatively safe path forward for the platform. However, not everyone inside Commodore believed that an evolutionary upgrade would be enough. Some engineers felt the industry was moving too quickly for incremental improvements to keep the Amiga competitive. Real-time 3D graphics were beginning to emerge, and new hardware architectures were appearing that could handle these tasks far more efficiently than the Amiga’s traditional design. These engineers began exploring a more radical approach that eventually became known as the Hombre project. Hombre represented a complete break from the past. Instead of building on the existing architecture, the project proposed an entirely new system built around modern technologies. The most striking change was the decision to move away from Motorola processors in favor of a RISC architecture. Early concepts suggested using Hewlett-Packard’s PA-RISC technology, which offered higher clock speeds and a simpler instruction set that could scale more effectively in the future.

The graphics system planned for Hombre was equally ambitious. Rather than focusing primarily on 2D graphics like previous Amiga chipsets, Hombre would incorporate hardware designed specifically for 3D rendering. Features under discussion included texture mapping, Gouraud shading, Z-buffering and perspective-correct rendering. These techniques were becoming essential for modern 3D graphics and were already appearing in high-end workstations. Hombre would also adopt chunky pixel graphics, a format far better suited for manipulating individual pixels and textures. This change alone would have made many modern graphics algorithms far easier to implement. The system architecture would move toward more standardized designs as well, incorporating a 64-bit memory interface and expansion buses similar to the PCI architecture that was beginning to appear in PCs. In many ways Hombre anticipated the direction that graphics hardware would take in the following decade. Dedicated graphics pipelines, texture mapping and specialized instructions for rendering would later become standard features of modern GPUs. At the time, however, these ideas were still cutting-edge and somewhat risky.

The biggest challenge facing Hombre was compatibility. Because the architecture was completely different from previous Amiga systems, existing software would not run natively. This was a serious problem because the Amiga’s large library of games and applications was one of its greatest strengths. Some engineers suggested including a small AGA-compatible subsystem inside Hombre machines so that older software could run in a compatibility mode. While technically possible, this approach would add complexity and cost. Ultimately the debate between AAA and Hombre reflected a deeper question about the future of the platform. Should the Amiga continue evolving its unique architecture, preserving compatibility and building on its existing ecosystem? Or should it leap into an entirely new design aimed at competing with the next generation of graphics hardware? Unfortunately Commodore never had the chance to answer that question. Throughout the early 1990s the company was struggling financially. Management problems, inconsistent marketing and increasing competition from PCs and game consoles all contributed to declining revenues. Engineering teams continued working on new technologies, but funding and resources were becoming increasingly limited.

AAA development experienced repeated delays as engineers attempted to solve technical challenges with limited budgets. Prototype chips were produced, but the project never reached the stage where it could be manufactured in commercial systems. Meanwhile Hombre remained largely conceptual, with early design work underway but no completed hardware. In April 1994 Commodore declared bankruptcy, becaused of the failed Amiga CD32 launch (legal battle for patents). With the collapse of the company, development of both AAA and Hombre came to an abrupt end. The ambitious plans for the next generation of Amiga hardware vanished almost overnight. Looking back today, it is fascinating to imagine what might have happened if these projects had reached completion. An AAA-based Amiga might have extended the life of the classic architecture, keeping the platform competitive in areas such as video production, animation and multimedia computing. A successful Hombre system, on the other hand, might have transformed the Amiga into a powerful 3D graphics machine capable of competing with early consoles and graphics workstations.

Although neither project reached the market, they remain important chapters in the history of the Amiga. They show how engineers inside Commodore were thinking about the future of computing and how they attempted to adapt one of the most innovative platforms of the 80s to a rapidly changing technological landscape. In the end AAA and Hombre represent two alternate futures that never came to pass. One path would have continued the legacy of the original Amiga architecture, refining and extending its capabilities. The other would have reinvented the platform entirely, anticipating many of the technologies that would later define modern graphics hardware. Both visions reveal the creativity and ambition of the engineers who worked on them, and both remind us that the history of computing is filled not only with successes but also with remarkable ideas that never had the chance to become reality.

Spread the love

MORE AMIGA & RETRO NEWS