There's a common misconception that to even dip your toes into the AI waters, you need to invest in state-of-the-art, high-end hardware. It's easy to understand why people believe this --- after all, until recently, you did need more robust hardware to run most models.
Just look at Stable Diffusion, a popular open source AI art-generation application. When it was initially released in August 2022, Stable Diffusion needed 16GB of RAM and an Nvidia graphics card with at least 10GB of VRAM. At that time in 2022, the Nvidia RTX 3080 card cost between $900 and $1300.
Now, however, less than two years later, Stable Diffusion's minimum system requirements are a graphics card with just 4GB of VRAM and can run on Windows, MacOS, or Linux operating system. You can pick up an Nvidia Quadro graphics card with 4GB of VRAM on Amazon today for less than $200.
As the landscape of AI hardware and software continues to evolve, even systems that require high-end hardware today are being optimized to run on more accessible hardware. This ongoing trend is making AI more inclusive and accessible to a broader audience.
That's why it's time to just say "no" to expensive, specialized hardware, especially for new users. You don't need fancy, super-advanced hardware to start your AI journey. You don't need the latest and greatest GPUs like the RTX 4090, and you certainly don't need a $10,000 gaming system. In fact, you can get started with AI using surprisingly modest hardware, such as a six-year-old laptop or an M1 Mac.
Many recent AI applications designed for individual use can operate just fine on hardware that meets the generic minimum requirements required for AI, including:
A Mac from the last 4 years. The M1 is Apple's first stab at a custom chip for their Macs that integrates the CPU and GPU, RAM, and other components on one chip. The M1 chip launched in late 2020 and delivers high performance and optimizations for a lot of workflows, including AI.
A computer with an Nvidia GTX 1050 graphics card or better. Launched in late 2016, the GTX 1050 card has a graphics processor that is suitable for gaming and AI, affordable, and found on many lower-cost new and used gaming laptops or desktops.
Either of these can be found easily and cheaply. Even thrift shops now have these systems and often price them below $400 with the cheaper options even going below $100.
Note that an M1 Mac and a GTX 1050 GPU are considered entry-level for AI applications. If you're going to work with high-end graphics or use more complex AI models, you may want a more robust computer. But if you're just getting started, this kind of hardware is more than capable of handling various AI projects, including our own Deepmake software which gives you access to all sorts of AI effects for VFX.
The takeaway here is this: AI technology is now within reach for everyone. You can start playing with AI today and run many models with the hardware you already have.
This democratization of AI is a significant step forward and makes it possible for enthusiasts and beginners to explore and experiment. Embrace the affordability and accessibility of AI hardware, and let your creativity flourish in the world of artificial intelligence. The important thing is to just get started.