Researchers at MIT's Computer Science and Artificial Intelligence
Laboratory (CSAIL) have developed a programming language called Halide for the implementation of multicore image
Not only are Halide programs easier to read, write and revise than
image-processing programs written in a conventional language, but because Halide
automates code-optimization procedures that would ordinarily take hours to
perform by hand, they're also significantly faster, says the team.
the MIT researchers used Halide to rewrite several common image-processing
algorithms whose performance had already been optimized by seasoned programmers.
The Halide versions were typically about one-third as long but offered
significant performance gains; two-, three-, or even six-times speedups. In one
instance, the Halide program was actually longer than the original but the
speedup was 70-fold.
However, the development is currently separate to the
OpenCL multicore programming specification.
Jonathan Ragan-Kelley, a graduate
student in the Department of Electrical Engineering and Computer Science (EECS),
and Andrew Adams, a CSAIL postdoc, led the development of Halide, and they've
released the code online.
Halide doesn't spare the programmer from thinking
about how to parallelize efficiently on particular machines, but it splits that
problem off from the description of the image-processing algorithms. A Halide
program has two sections: one for the algorithms, and one for the processing
"schedule." The schedule can specify the size and shape of the image chunks that
each core needs to process at each step in the pipeline, and it can specify data
dependencies — for instance, that steps being executed on particular cores will
need access to the results of previous steps on different cores. Once the
schedule is drawn up, however, Halide handles all the accounting
A programmer who wants to export a program to a different
machine just changes the schedule, not the algorithm description. A programmer
who wants to add a new processing step to the pipeline just plugs in a
description of the new procedure, without having to modify the existing ones. (A
new step in the pipeline will require a corresponding specification in the
"When you have the idea that you might want to
parallelize something a certain way or use stages a certain way, when writing
that manually, it's really hard to express that idea correctly," said
Ragan-Kelley. "If you have a new optimization idea that you want to apply,
chances are you're going to spend three days debugging it because you've broken
it in the process. With this, you change one line that expresses that idea, and
it synthesizes the correct thing."
Although Halide programs are simpler to
write and to read than ordinary image-processing programs, because the
scheduling is handled automatically, they still frequently offer performance
gains over even the most carefully hand-engineered code. Moreover, Halide code
is so easy to modify that programmers could simply experiment with half-baked
ideas to see if they improve performance.
"You can just flail around and try
different things at random, and you'll often find something really good," said
Adams. "Only much later, when you've thought about it very hard, will you figure
out why it's good."This article was first posted on our sister site EE Times Europe.
Related links and articles:
Mali cores support Khronos compression
ARM supports European research on GPU programming
Altera's OpenCL for FPGAs dramatically reduces development times