Chris Kranky

Recent Posts


The real story on VP8 and H.264

Chris KoehnckeChris Koehncke

soze2Is the debate for WebRTC on VP8 and H.264 settled? An why was there even a debate? Industry pundits wrote about which one was technically superior (neither or both or pick one – all are valid answers).  But no one really explored the single most important issue at hand — money. But I have. An all roads lead back to one person. Keyser Soze.

Keyser Soze in this case, isn’t a person, it’s a company. But like Keyzer Soze, they profess to be innocent and uninvolved while all kinds of bad things happen around them for their profit. A company who operates with a net profit of 30% (minting money into the billions), they are armed with the most dangerous weapon on the face of the plant, a library of patents and buildings full of lawyers. Meet Keyzer Soze, meet Qualcomm.

If you’ve never heard of Qualcomm, they’re just as happy for it. In fact, 99% of Qualcomm’s high profitable revenues are from outside the US, indeed 80% of their revenues come from just 3 countries and a mere 1% comes from it’s home base in the US. What 3 countries you ask? That’s easy. China, Korea and Taiwan.

According to Qualcomm, they play a leading role in:

In other words Qualcomm patents at the very heart of all things related to mobile electronics and loads of areas where WebRTC is playing about. Their lawyers have been extremely busy ensuring that their patents are enforced and valid in the emerging economies of Brazil, India and China. The why apparent in a moment.

So what about the H.264? Google could care less about H.264 and it’s licensing costs. Google likely has already enough valid H.264 licenses sitting in someone’s desk drawer to cover their needs for the next 10 years. That’s not the battle.

The battle is mobile and in particular smartphone mobile and most specifically smart phone adoption in emerging countries.

Costs matter and a $1 of licensing cost easily translates to a $3-4 in sales price. Hiding a $1 of licensing cost into a $700 smart phone is pretty easy. Doing it in a $70 smart phone is more difficult. India has a median income of $616 PER YEAR. Brazil is only slightly better at $938 and China is downright rich at $10k per year (America is $52k by the way). Phones will need to be cheap to win in these markets. Hint: Cheaper phones in India will also mean cheaper phones in today’s lucrative Western markets. The $700 iPhone will start to look downright expensive as the “more for less” axiom grinds away.

It become clearer now what the battle is.

Inexpensive smart phones will be needed to drive e-commerce and advertising in the growth emerging countries. Simply put, the cheaper the phone, the more than people who can buy afford one with a result of more people you can sell stuff to. To make an inexpensive smart phone, you have to drive the underlying costs out. As you create your bill of materials you quickly discover a cost item that you have no ability to cost reduce, negotiate or otherwise eliminate. That item is named “Qualcomm Licenses”.

Thus I offer that the H.264 vs VP8 argument spans far beyond a video codec technology choice or even perhaps a simple H.264 license. Instead, H.264 represents the tip of the iceberg in a mountain of non-negotiable licensing costs associated with mobile devices. Google is likely working on numerous fronts to open source, acquire and give away as much underlying mobile technology as they can. Cheaper smart phones directly benefit them in billions of ways.

I’d fathom the discussion of H.264’s inclusion into WebRTC is far from over as this game of Stratego continues on various fronts but keep your eyes on mobile.