Advertisement

New computer program may fix billion-dollar bit rot problem

"We've found that Helium can make updates in one day that would take human engineers upwards of three months," said researcher Saman Amarasinghe.

By Brooks Hays

BOSTON, July 9 (UPI) -- Bit rot is a billion-dollar problem for software developers. But now computer engineers may have designed the ideal code-fixing program to combat it.

High performance computer programs, like Photoshop and other image editing software, have to be optimized for the hardware environment they exist within. But hardware and the program's technological surroundings are constantly changing, so software must be continually re-optimized.

Advertisement

The result is increasingly tangled hunks of old code, which require more and more manpower each time program iterations need to be updated and optimized once again. The time-intensive and burdensome process is expensive for software companies like Adobe and others.

That is why Adobe and other developers are so excited about the prospects of Helium, a computer program which promises to smooth out old code more efficiently (and more cheaply) than human engineers.

The program, which was created by students at MIT, is designed to scan for the most significant computational components -- binary codes called "stencil kernels" that serve as building blocks for more complicated algorithms. Once located, the old and rotted kernels are replaced with newly optimized components.

"The order of operations in these optimized binaries are complicated, which means that they can be hard to disentangle," Charith Mendis, a graduate student at MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), explained in a press release. "Because stencils do the same computation over and over again, we are able to accumulate enough data to recover the original algorithms."

Advertisement

Mendis is the lead author of a new paper describing Helium's abilities.

"We've found that Helium can make updates in one day that would take human engineers upwards of three months," added Saman Amarasinghe, a professor at MIT and researcher at CSAIL. "A system like this can help companies make sure that the next generation of code is faster, and save them the trouble of putting 100 people on these sorts of problems."

In other words, even computer engineers may soon their jobs to computers.

Latest Headlines