NVIDIA Digits: can it be the new “Amiga”? | Festina Lente - Your leading source of AI news | Turtles AI

In 1985, Commodore introduced the Amiga 1000. It was a machine ahead of its time. History, as they say, often repeats itself. Now, 40 years later, NVIDIA has unveiled Project Digits. The parallels between these groundbreaking systems are striking—let’s explore them together!

Commodore Amiga

The original Amiga 1000 featured a Motorola 68000 processor and 256KB of RAM (that soon became 512KB and 1MB in later models). Its custom chipset enabled advanced graphics and sound capabilities. This made it a favorite among gamers and creative professionals.

The Amiga’s preemptive multitasking operating system, AmigaOS, was revolutionary. It allowed users to run multiple applications simultaneously without system crashes. This was a significant advancement over competitors like the IBM PCs of those times.

The Amiga’s graphics capabilities were outstanding. It could display up to 4096 colors simultaneously using Hold-And-Modify (HAM) mode. This was a feat that left other systems in the dust. The sound capabilities were equally impressive. The Amiga featured four-channel stereo sound, providing rich audio experiences both for games and professional applications. Its hardware design included a blitter chip (Block Image Transfer), which accelerated graphics by handling tasks like bit blitting, shape drawing, and layering directly in hardware, reducing CPU load. Also, its custom designed chips included Agnus (managing memory and display rendering), Denise (handling advanced graphics like sprites and HAM mode), and Paula (providing four-channel stereo sound and managing I/O). All this was unprecedented, at least in computers of this class

Amiga’s design was modular. Users could expand its capabilities with peripherals and upgrades. This flexibility contributed to its popularity among enthusiasts.

The gaming community embraced the Amiga. Its superior graphics and sound made it a preferred platform for game development. Titles like ”Shadow of the Beast”, ”Another World”, "Lemmings", "The Secret of Monkey Island" and many others, showcased the Amiga’s capabilities. In the field of video production, the Amiga found a niche. Its genlock capabilities allowed for seamless video overlay, a feature valued by broadcasters. The Video Toaster, an add-on for the Amiga, became a popular tool in television production. It enabled affordable video editing and effects for both professionals and enthusiasts, democratizing the industry (most visual effects of Babylon 5 were realized in this way).

NVIDIA Digits

Fast forward to 2025. NVIDIA unveils Project Digits, a “personal AI supercomputer”. Priced at $3,000, it aims to democratize AI development. At its core is the GB10 Grace Blackwell Superchip. This chip delivers up to 1 petaflop of AI performance at FP4 precision. Project Digits comes with 128GB of unified memory and up to 4TB of NVMe storage. It can handle AI models with up to 200 billion parameters. For more demanding tasks, two units can be linked to manage models up to 405 billion parameters. The system supports popular frameworks like PyTorch, Python, and Jupyter notebooks. It runs on a Linux-based operating system, NVIDIA DGX OS. NVIDIA’s collaboration with MediaTek optimized the GB10’s power efficiency. This partnership mirrors past collaborations that pushed technological boundaries. Project Digits integrates with NVIDIA’s AI software library. This provides developers with tools for experimentation and prototyping. The system’s compact design allows it to fit on a desk and operate from a standard power outlet. This accessibility is reminiscent of the Amiga’s appeal to individual users and small developers. NVIDIA’s move into personal AI computing challenges existing market players and established architectures, like of course x86. Companies like Intel and AMD may feel the heat from this new competition. The pricing strategy positions Project Digits as an accessible tool for a wide range of power-users. This approach echoes the Amiga’s initial market positioning for the Amiga 1000 and Amiga 2000.

The potential applications for Project Digits are vast. From AI research to data science, it opens doors for innovation. Its release may inspire a new wave of software development tailored to its capabilities. The AI community has responded with enthusiasm. Discussions highlight the system’s potential to revolutionize local AI development.

A new "Amiga"?

The parallels between the Amiga and Project Digits are striking to me. Both introduced advanced technology aimed at empowering individual users. Each entered a competitive landscape with something radically different from anything else available on the market at the time. Both were built on powerful custom hardware for their era, capable of potentially sparking new applications, while competing with much more expensive professional solutions.

In particular, both systems aim to empower individual users by providing them with cutting-edge tools to engage deeply with the most transformative technologies of their respective eras: for the Amiga, this meant enabling personal video and audio production through groundbreaking features like genlock, which revolutionized video overlay and made creative media production accessible to enthusiasts; for Project Digits, it signifies granting unprecedented access to advanced AI model training and experimentation, allowing individuals to harness the power of machine learning and artificial intelligence in ways that were previously confined to large corporations or research institutions able to afford expensive boards like H100.

Just as the Amiga inspired the birth of the demo scene, where independent developers showcased audiovisual mastery through programming, Project Digits could foster a similar culture for AI. With its accessible price point, Digits might empower independent developers and small studios who cannot afford larger systems like the DGX ones, creating a vibrant community of experimental AI projects.

Digits also holds the potential to cultivate a robust developer community, much like the Amiga did in its era. By providing tools for open experimentation and embracing popular frameworks like PyTorch and Python, this new machine might inspire a thriving ’scene’ where developers and hobbyists share custom models, innovative applications, and AI-driven tools, building a grassroots ecosystem of collaboration.

Additionally, Digits could become a powerful educational tool, democratizing access to AI education in ways the Amiga once did for graphics, sound and programming. Schools and universities might adopt the system to teach students advanced concepts in machine learning, allowing hands-on experimentation with models and data.

The story of the Amiga offers lessons. Technological superiority alone doesn’t guarantee enduring success. Market strategy, user adoption, and ecosystem support are equally important.

The Amiga’s downfall in the early 90s was partly due to poor marketing and limited software support. NVIDIA’s established brand and extensive software ecosystem may help Project Digits avoid a similar fate. The success of Project Digits will depend on several factors. Market adoption, software development, and competitive responses will play crucial roles. The computing landscape has evolved since the Amiga’s time. Today’s market is more interconnected and rapidly changing. Project Digits enters a world where AI is a focal point of technological advancement. Its impact could be profound, influencing both hardware and software development.

As Project Digits prepares for its May release, the tech community watches closely. Will it carve out a small niche, or will it redefine personal computing?

What if...Gaming Meets AI?

Now, let me play with my imagination. What if NVIDIA decided to create a "budget" version of Digits, perhaps with 32GB of memory instead of 128GB and slightly reduced specifications? And what if this version sparked the development of video games specifically tailored for it? And what if it were integrated with a keyboard? Welcome, new Amiga 500.

Interestingly, a lesser version of the GB10 chip could even become the foundation for future gaming consoles, much like how the Amiga’s architecture influenced systems like the Atari Lynx, 3DO and, indirectly, even the Playstation with its RISC CPU and its GTE and SPU custom coprocessors. By embedding AI-driven capabilities into consoles (like small language models), these devices could redefine real-time interactivity and personalized gaming experiences. After all, the GB10 is based on Blackwell, that also powers the new 50x0 Nvidia cards.