# Changeset 1778 for docs/HPCA2012/final_ieee

Ignore:
Timestamp:
Dec 13, 2011, 9:24:59 PM (8 years ago)
Message:

Minor fixes.

Location:
docs/HPCA2012/final_ieee
Files:
2 edited

### Legend:

Unmodified
 r1775 This overhead has little impact on the overall performance of these parsers as they experience additional overheads related to branch mispredictions. Wne compared with Xerces and Expat, the data organization of Parabix-XML When compared with Xerces and Expat, the data organization of Parabix-XML significantly reduces the overall cache miss rate; specifically, there were $7\times$ and $15\times$ fewer L1 and L2 cache misses compared to The high branch misprediction rate of conventional parsers is a significant overhead, with the cost of a single branch mispredic- tion on the order of 10s of CPU cycles spent to restart the processor significant overhead, with the cost of a single branch misprediction on the order of 10s of CPU cycles spent to restart the processor pipeline. Parabix-XML is nearly branch free and exhibits minimal dependence on the source markup density. Specifically, our study