It’s that time of year again – SIGGRAPH is here! For computer graphics artists, teachers, freaks and geeks of all descriptions, it’s like having Midsummer, Christmas, and your birthday all in the same week. By the time you read this, I’ll be in beautiful Vancouver BC, happily soaking up the latest in graphics research, technology, animation, and associated general weirdness along with the other 15,000-plus attendees. I can’t wait!
This year, SIGGRAPH has a special personal connection for me: my office-mate Dave Shreiner is this year’s general chair (amazingly, he’s still got all his hair – quite a lot of it actually), and my other office-mate Jesse Barker is chair of SIGGRAPH Mobile. (Jesse’s got no hair at all, but with him it’s a style choice.) My own job at SIGGRAPH is a lot less grand, but it’s something I love doing: In my capacity as OpenGL® ES working group chair, I’ll be co-hosting the Khronos OpenGL / OpenGL ES Birds of a Feather (BOF) session. That’s where the working groups report back to the user community about what’s going on in the ecosystem, what the committee has been doing, and what the future might hold. This year’s OpenGL ES update will mostly focus on the growing market presence of OpenGL ES 3.0, and on OpenGL ES 3.1, which we released earlier this year and which is starting to enter the market in a big way. It’s great stuff – but it’s not the big news.
There’s a change coming
By the standards of, well, standards, the OpenGL APIs have been an amazing success. OpenGL has stood unchallenged for twenty years a cross-platform 3D API. Its mobile cousin, OpenGL ES, has grown phenomenally over the past ten years; with the mobile industry now shipping a billion and a half OpenGL ES devices per year, it has become the main driver of OpenGL adoption. One-point-five billion is a mind-boggling number, and we’re suitably humbled by the responsibility it implies. But the APIs are not without problems: the programming model they present is frankly archaic, they have trouble taking advantage of multicore CPUs, they are needlessly complex, and there is far too much variability between implementations. Even highly skilled programmers find it frustrating trying to get predictable performance out of them. To some extent, OpenGL is a victim of its own success – I doubt that there are many APIs that have been evolving for twenty years without accumulating some pretty ugly baggage. But that doesn't change the central fact: OpenGL needs to change.
The Khronos working groups have known this for a long time; top developers (hi Rich!) have been telling us every chance they get. But now, with OpenGL ES 3.1 finished but still early in its adoption cycle, we finally feel like we have an opportunity to do something about it. So at this year’s SIGGRAPH, Khronos is announcing the Next Generation OpenGL initiative, a project to redesign OpenGL along modern lines. The new API will be leaner and meaner, multicore and multithread-friendly. It will give applications much greater control over CPU and GPU workloads, making it easier to write performance-portable code. The work has already started, and we’re making rapid progress, thanks to strong commitment and active participation from the whole industry, including several of the world's top game engine companies.
Needless to say, ARM is fully behind this new direction, and we’re investing significant engineering resources in making sure it meets its goals and runs well on our Mali GPUs. We are of course also continuing to invest in the ecosystem for ‘traditional’ OpenGL ES, which will remain the dominant mobile graphics API for quite some time to come.
That’s all I’ve got for now. If you’re going to be at SIGGRAPH, I hope you’ll come by the OpenGL / OpenGL ES BOF and after-party, 5-7pm on Wednesday at the Marriott Pinnacle, and say hi. If not, drop me a line below…
Tom Olson is Director of Graphics Research at ARM. After a couple of years as a musician (which he doesn't talk about), and a couple more designing digital logic for satellites, he earned a PhD and became a computer vision researcher. Around 2001 he saw the coming tidal wave of demand for graphics on mobile devices, and switched his research area to graphics. He spends his working days thinking about what ARM GPUs will be used for in 2016 and beyond. In his spare time, he chairs the Khronos OpenGL ES Working Group.