|
Efficiency of Markov chain implementations Execution time for a 31-state GHMM gene finder utilizing hash tables or FSMs for its content sensors, applied to a 1.8 Mb sequence. The FSM implementation was over twice as fast as the hash table implementation, and required significantly less memory. |
||
| time (min:sec) |
total RAM |
|
|
|
||
| DSP/Hash |
1:15 |
53 Mb |
| DSP/FSM |
0:34 |
44 Mb |
Majoros et al. BMC Bioinformatics 2005 6:16 doi:10.1186/1471-2105-6-16 |
||