Hacker News new | ask | show | jobs
by glhaynes 5142 days ago
I think you've misunderstood my point which was that Microsoft engineers are very capable of reducing memory usage if charged to do so (and, as you say, history bears this out) and that I would expect them to likewise be able to reduce power usage [within the constraints provided by environment and hardware requirements].
2 comments

There's a big difference between reducing memory pressure/performance (which kind of solved itself once applications stopped increasing memory usage faster than the price/Mb dropped, and the move from 32-bit to 64-bit helped a bit too; handling stupid things like XP swapping everything out to disk overnight were easy wins as well) and improving power efficiency (requires application and OS redesign, which Microsoft have done and judging by the moaning about WinRT in this thread by developers the uptake will be a challenge; even then if you improve efficiency people just keep using the damn devices even more so it's really difficult to win whereas with memory there's a typical usage value that changes safely relative to RAM increases as applications grow over the years, whereas battery capacity isn't improving as rapidly).

I'm throwing another 12GB of RAM (making 24GB total) into my system tomorrow when I install Windows 8. Together with my new SSD it'll help crunching through some large data sets so that I can hit 100% CPU utilization. It cost me $100 USD to do so in a country where we often pay double for electronics. Few "normal" people can utilize all of that during the normal course of using a PC. Memory pressure is an easy problem to solve.

Memory is cheap. There's probably better things to focus on then squeezing more bits out of the memory usage.