Where's the Real Bottleneck in Scientific Computing?
Scientists would do well to pick up some tools widely used in the software industry
On the Hard Drives of Giants
Science is much more than just a body of knowledge. It's a way of doing things that lets people separated by oceans, decades, languages and ideologies build on one another's discoveries. Computers are playing an ever-larger role in research with every passing year, but few scientific programs meet the methodological standards that pioneers like Lavoisier and Faraday set for experimental science more than 200 years ago.
Better education is obviously key to closing this gap, but it won't be enough on its own. Journals need to start insisting that scientists' computational work meet the same quality and reproducibility standards as their laboratory work. At the same time, we urgently need more journals willing to publish descriptions of how scientists develop software, and how that software functions. Faster chips and more sophisticated algorithms aren't enough—if we really want computational science to come into its own, we have to tackle the bottleneck between our ears.