Rico Mariani's Performance Tidbits

Raymond Chen is running a series of articles about how to build and optimize the startup time of a Chinese/English dictionary. Developing a Chinese/English dictionary: Introduction Loading the dictionary, part 1: Starting point and then my analysis...

Well, I've been slacking off too long... sitting on my nice 124ms time and doing nothing. It's time I did something to the managed code. You'll recall I had some observations in my analysis of the managed code and basically I can summarize my thoughts...

My last entry was some generic advice about how to do a good performance investigation. I think actually it's too generic to be really useful -- in fact I think it fails my Peanut Butter Sandwich Test. Digression to discuss the Peanut Butter Sandwich...

Well, it's time for me to surrender. Sort of :) Raymond pulls out all the stops in his sixth version by painting a big bullseye on his biggest remaining source of slowness which is operator new. He turns in an excellent result here. On my benchmark...

Raymond didn't post an update today so I think this is a good time for me to post the first analysis of the managed code. There's some things to notice right away and I've highlighted them in the table below. This table was generated in the same way as...

Yesterday Raymond posted his initial version and I posted Performance Quiz #6 . Today he posts the first optimization . Quoting him: Upon profiling our dictionary-loader, I discovered that 80% of the CPU time was spent in getline . Clearly this...

I find that sometimes people have difficultly just getting started when doing a performance analysis – meaning they’re faced with a potentially big problem and don’t know where to begin. Over the years many people have come to me under those circumstances...

The fun continues as today we look at Raymond's third improvement . Raymond starts using some pretty hefty Win32 magic with a mapped file view to eliminate once-and-for-all any costs associated with the getline function. And good news, his new version...

Raymond is definately making some great headway. Having targetted the single-character at a time conversion problem in version 4 of his program he's 1.84 times faster than the previous version. Version Execution Time (seconds) Unmanaged...

Stefang jumped into the fray with his analysis in the comments from my last posting . Thank you Stefang. And it seems, as usual, things aren't always what they appear :) Let's continue the story without me giving away the ending just yet though. How...

I thought I would end this series by looking at the space costs of both solutions because it gives a real window into what I do every day and perhaps it's helpful to understand the multidimensional aspects of performance analysis. It's also interesting...

Well today Raymond hits a snag in version 5 of the program. I started by profiling the code as provided (in fact I had the .exe from Raymond himself so I didn't even have to build it). The new version is actually slower (start to finish) than the previous...