I'm Morgan McGuire (@CasualEffects). I've been working on computer graphics and games for 20 years at great places including NVIDIA, Williams College, Brown University, and Activision.

See my home page for a full index of my blog posts, books, research, and projects.

Friday, February 6, 2015

A Billion in Computer Graphics

The Bell Labs Transistor

The Transistor

In 1947, there was one transistor on the planet. It looked like a high school science project and could just barely fit in a large pocket. The research leading to it was expensive. The transistor itself was also expensive because it was made out of gold.

A decade of refinement later, consumers encountered their first transistor product. A new brand called Sony created a portable radio. The radio contained four transistors, fit in a large pocket, and cost the equivalent of $350 in 2015 dollars. This was not an instant success, but the brand and the technology lived on.

Today, there are around four billion transistors in the smartphone in your pocket. The material cost of a modern transistor-based processor is close to zero because the main ingredient is sand.

Between the first consumer product with transistors and today's smartphone is an engineering factor of one billion: 1,000,000,000. That's the scale difference between holding a dinner party and feeding the world's population.

In 1947, Bell Labs invisibly lit the fuse of the most revolutionary change in history. By 1957, the first effects were public but widely judged a curiosity. Very few people who looked at the expensive, awkward entertainment device of a four-transistor radio a half-century ago appreciated its significance. Yet by 1997, transportation, entertainment, finance, education, medicine, warfare, and communication in the industrialized world were completely dependent on transistors. As a result, today's technologies are as different from their 1947 counterparts as magic.

Virtual Reality

Samsung Gear VR
In 2015, the first virtual and augmented reality head-mounted displays for consumers will hit the market. They'll cost about $350 and will be bulky. Oculus VR (owned by Facebook), Microsoft, Sony, HTC, Zeiss, LG, and Samsung have announced such products and there are likely many unannounced competitors such as Magic Leap.

We'll also see personal hand, body, and eye tracking devices from companies such as Tobii, Microsoft, and Leap Motion. These will provide input that matches the immersive output of the head-mounted displays. All of this technology has been incubating in labs, looking like high school science projects. The first consumer products will be only slightly better, built from crudely repurposed waveguides, lenses, projectors, light panels, and micro cameras.

The coming first generation of augmented and virtual reality devices contains niche entertainment products. They're hard to use and not nearly as compelling as their advertising and a background in science fiction films leads you to believe. They may be commercially successful, or they may fail in the market at first and retreat to the labs for another two decades.

Investors see virtual reality as a gamble with potential for a high reward. You can read their opinions in any business newspaper, and it isn't a strong indicator. An important sign which newspapers haven't reported is a sea change across the scientists and engineers in computer graphics. Many, including the most accomplished and brightest, have recently disappeared from public view. They've changed employers or research areas and now work on virtual reality. Why?

Most people look at a 2015 virtual reality headset and see that it is expensive, underwhelming, and awkward. They see a four-transistor radio. We look, see it advance a billion times, and leap toward future history.



Morgan McGuire (@morgan3d) is a professor of Computer Science at Williams College, a researcher at NVIDIA, and a professional game developer. He is the author of the Graphics Codex, an essential reference for computer graphics now available in iOS and Web Editions.