Simply beyond ridiculous (H265 proposal testing)
x264dev.multimedia.cx
x264dev.multimedia.cx
But I can't say for sure that there weren't some totally insane encoders put forward by those involved.
What I suspect is going on is a case of exponents. That is, you have an algorithm that tries every possible A, an algorithm that tries every possible B, and an algorithm that tries every possible C. You optimize them independently, and then in the final encoder you submit, you run them all at the same time, optimizing every possible A, for every possible B, for every possible C, sending runtime sky-high.
I'd read the code, but it's nearly-undocumented C++, written by the kind of people who reimplemented their own option parser using C strings (in C++) in Visual Studio.
The author didn't consider the use of GPU to perform the encoding. Kind of fishy. For my application (not video encoding) I managed to reduce the computation time from 5h to 15s (600x faster!) by using a GTX280 GPU card.
More importantly, regardless of how important you think encoding is, if I can't encode a 5 second test sequence in less than a day, I'm not going to be able to do much experimentation with the software. Sure, you can port it to a GPU, but right now it isn't on a GPU, and right now, I'd like to experiment with it.
If the algorithms have a lot of data reuse in their matrix computations, I can see achieving 600x improvement when compared to a hardware cache based architecture. If the CPU implementation doesn't do tiling (http://en.wikipedia.org/wiki/Loop_tiling) effectively (or it can't) then it's going to shuttle a lot of data back and forth from cache to RAM.
The accepted term for this effect is super-linear speedup.
The spec is being written now. The encoder needs to be tested and evaluated now. There's no point in doing it now if we'll have to wait 10 years until CPUs can be used for testing whatever you make.
> Also, that software was likely to have a lot of debugging flags, etc.
It was released in source form, so those are easily disabled / removed.
(this is a bit of a red herring - the times related in the story are pretty much useless)
The GPU, looking at each individual core, is actually a very weak general purpose processor. GPGPU is good for because there are a lot of cores, so you can run highly parallelizable tasks easily in it.
Video encoding is not one of these tasks since each block depends on the previous block, each frame depends on the previous, and even on the next frame.
There have been uncountable proposals to port or accelerate a part of x264 on GPGPU. Nobody has succeeded in two years, not even for the motion search, which is supposedly the component that would be the easiest to port and benefit the most from GPGPU.
I expect that video encoding has the same pattern. One of the critical aspect to benefit from GPU parallelism is the amount of state information each thread has to maintain. It has to be kept to a minimum because this space is limited. If they need more space the number of active threads is reduced to match the requirement.
At first I read this and couldn't believe it. A 40% improvement! Wow!
But reading the rest of the article, it's disappointing to see that it's not a practical %40...
Makes me wonder if GPUs could be used to do the encode and speed it up enough to make it practical (maybe within a few generations of GPUs..). After all, you have to aim for where computers will be by the time this becomes a standard, not where they are now.
Does using a GPU really help as much as chmike suggests?
What do the encoder's authors have to say?
Would GOOG fast-track vp8 as an ISO standard, like MS did with OOXML? After all, it is ubiquitous in flash video, just as Word is -- a de facto standard needing rubber stamping.
Also note that Microsoft only got to fast-track OOXML as an ISO standard because it had already been passed as a standard by ECMA, not because of anything inherent in OOXML (which wasn't even a de facto standard, the binary .doc was).