Faults, reclaim, IPI's, system daemons: They all take away the processor from the code that the user wants to run. In many areas the processing latencies requirements are now calcualted in microseconds. The indeterminism caused by memory management becomes a significant headache for HPC apps that need to rendevouz in deterministic intervals or for HFT apps that need to complete a trade ASAP. Isnt there a way that we can get the OS out of the hair of the applications? Maybe only temporary so that the app can process undisturbed for some time? Or is it possible to shift processing to a dedicated processor that is dedicated to the OS? This becomes more severe each year as the complexity of memory management and therefore the overhead created by OS grows. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>