|P is for Practical
It is indeed possible to speed up floating point conversion. I actually read a blog article (with decent links to further information) on recent advances on the subject; see here.
However, I seriously doubt this is truly the bottleneck in your program. Have you profiled? Are you asking for improvement in CPU speed, or perhaps in programmer time spent using and understanding the conversion?
In reply to Re: fastes way to print floats with fixed decimal places