Skip to main content
Software Engineering

You are not logged in. Your edit will be placed in a queue until it is peer reviewed.

We welcome edits that make the post easier to understand and more valuable for readers. Because community members review edits, please try to make the post substantially better than how you found it, for example, by fixing grammar or adding additional resources and hyperlinks.

7
  • 1
    A problem is required to be (a) parallelisable and (b) have well-defined standards to be considered for or benefit from dedicated hardware (such as a GPU). Text search can be broken down into smaller parts as long as the target is large. Most other operations that you've mentioned can't - streaming the output is as difficult as the operation involved. Commented Dec 8, 2014 at 20:57
  • 2
    a good overview on FPGA's and their applicability to stream processing can be found here, which could be extended to text processing, I guess. However, text processing often involves frequency counting of words and phrases, as well as indexing, which I doubt is more efficient if done at the hardware level. Except if your inputs are known beforehand, which kind of defeats the purpose of text processing. Commented Dec 8, 2014 at 21:01
  • related (possibly a duplicate): Why don't computers come with specialized hardware such as sorting networks? Commented Dec 8, 2014 at 21:01
  • It already exists and it is already implemented in various forms in common desktop processors. String operations have been commonplace on computers since nearly day 1, and many architectures have had many different instructions catering to particular string handling schemes. Commented Dec 8, 2014 at 21:50
  • 3
    Just buy an IBM mainframe -- had nearly all of these string operations as basic instructions since 1966. Commented Dec 9, 2014 at 9:46