Tegra K1 – What’s it worth?

I hardly consider myself any sort of authority on anything. I’m just another dude with an opinion who makes games and teaches people to make games. Thats why I generally avoid commenting outside of my area of expertise – generally I stick to talking about events I’ve been to, teaching things, some game design related stuff. Things I’ve done, screwed up and found solutions for.

This post will be a little different. This post is a piece of commentary, which I haven’t really done before.

This week was CES2014 – Consumer Electronics Show – and there’s always some great announcements. This year the one that has caught everyone’s eye is the announcement of the Tegra K1 – a 192-core mobile chip based on Nvidia’s Kepler desktop graphics architecture.
This announcement is big for a few reasons, not the least of which being accompanying videos showing Unreal Engine 4 running on a K1-powered device, and the reveal that it was a DirectX 11 chip.

Many websites have covered the announcement and tech specs of the processor, so there’s no point repeating all that. What I want to talk about are the potential implications for mobile game developers.

As developers we’ve had fairly powerful chips for a while. Granted, those chips aren’t anywhere near the power of the K1 but we’ve been able to make some pretty sexy looking games as it is. So what opportunities does one chip in a saturated marketplace offer to developers?

Despite a massive increase in power and the promise of coming close to matching PC quality graphics, I still believe there will be challenges developing for mobile devices in the forseeable future. While we do have DirectX 11 support on the K1, that will surely be limited to just Windows mobile devices which only make up a small percentage of the mobile market. Developers working with DirectX on PC will still need to work on dual rendering pipelines in order for games to run on mobile devices.
The proliferation of these chips is another potential concern. We are yet to see what else will come of the next generation to compete with the K1 for power, and without exclusive partnerships from hardware & device manufacturers I feel that developers would hesitate to design a game that focuses purely on the high-end, as its such a small marketplace.
Finally, the last challenge is that mobile devices are still inherently casual devices. 2D or small simple titles still totally dominate the mobile marketplace and given the lack of success by high-quality AAA-style 3D games on mobile devices in the past, you have to wonder where these style of games fit into the mobile economy.

We’re still probably a few years away from games being developed/ported to mobile in the same way that PC-to-console works right now. For developers who plan on using high-end engines like Unreal Engine 4 to take advantage of the better chips there are associated risks, given the majority of any target market will be in the mid-range of devices come next generation. But that doesn’t mean its not a risk worth taking – the payoff for a quality mobile title can still be huge.

Its going to be a hard for higher-end mobile developers to strike that balance. Perhaps a good first step is even some sort of asset sharing between systems. What I do believe is that this that Nvidia may have fired the first shot in what will be a long battle for mobile chip producers.

Leave a Reply