What kind of cpu resources does fpga "compilation" (Analyse, Synthesis, etc..) use on a cpu..?
Integer/Branch/Bitshifting..? Floating point..? Will a pipeline cpu greatly improve speed..?
What kind of cpu resources does fpga "compilation" (Analyse, Synthesis, etc..) use on a cpu..?
Integer/Branch/Bitshifting..? Floating point..? Will a pipeline cpu greatly improve speed..?
I always assumed it was sorting and searching types of operations. But I have no references to back that up.
Perhaps faster memory or a larger cache would help. Aren't all processors pipelined nowadays?
Alan Nishioka
Sounds resonable. The tough task must be to figure out optimal placement of "gates" and how to route switches. And when to relocate and let signals through chained switches (kind of like leased lines in telecomm).
Maybe the task is very much similar to BGP (Border Gate Protocoll) in terms of algorithms.
Proberbly. But if the app is benefitting a lot from pipelineing. Then it might pay to use a cpu with extra long pipeline.
I'm sure there's no floating point, except to report % usages at the end. Pipelining always helps, but anything newer than a 286 has at least some level of pipelining. Plenty of main memory and cache will help greatly. Swapping to disk will kill this type of large database sorting task.
Jon
I have found that cache size and FSB speed/size to be more of a determining factor of the compilation speed than clock speed.
I would image this is because the data structures formed would get pretty large (and wouldneed to be frequently accessed).
-Isaac
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.