Hobbyist use of FPGAs is currently stuck in the 1980s. Happily, we are talking about the good part of the past -- not the cheesy talking cars of Knight Rider, but instead the bleeding edge of computing.
1983 was a golden year for geeks. Return of the Jedi had just appeared in the theaters. My all-time favorite arcade game, Crystal Castles, had just been released. Microsoft shipped Windows 1.0, while Apple released the "Lisa" -- the product that would become the Macintosh -- which cost $9,999 for a 5MHz CPU and 1MB of RAM. The movie Tron had also just been released (the director later said that he believed the reason this movie didn't get nominated for an Oscar for special effects was that "The Academy thought we cheated by using computers").
In fact, if you look really closely at Tron, you might catch a glimpse of a Cray Supercomputer. In 1983, the Cray X-MP was the world's fastest computer -- Dual CPUs, 16MB of RAM, and a peak performance of 400 Megaflops. It had everything -- even a built-in sofa!
The Cray X-MP came with a free sofa (the sofa actually contained the cooling system).
"So what has all this got to do with FPGAs?" I hear you say. Well, that level of computing is pretty much where low-end FPGAs are today. For example, my FPGA-based fractal viewer -- running on a Spartan 6E LX9 FPGA -- computes about 720 million 36-bit multiplications per second. Meanwhile, my Terasic DE0-nano has 32MB of RAM. In fact, Chris Fenton has implemented a Homebrew Cray 1A Supercomputer on his FPGA development board.
It's amazing when you come to think about it. Sitting on my (more comfortable than a Cray) sofa in the evening with my laptop computer, I can implement designs that would have required high-end research programs only 30 years ago. A Cray used to cost $15 million, but my Papilio Plus FPGA development platform costs less than $100 -- including shipping!
I personally think this is fantastic, but I'm sure some people would say “Why would I want to bother recreating 80s technology at all?” In reality, unless you have a passion for something like old arcade games, you probably don’t. Instead, you want to do something new and exciting -- maybe something like:
Without FPGAs, projects of this caliber would be out of reach for even small commercial laboratories. As hobbyists, we may have gained access to bleeding-edge 1980s technology, but that's no reason we can’t party like it’s 1999!
Tweet This
