Multicores
|
From: robertwessel2@yahoo.com on 7 Sep 2005 16:56 Jan Vorbrüggen wrote: > > Different versions of Windows use different time-slice intervals. And > > there are some semi-documented ways that you can change that. Windows > > takes considerable effort to keep the system responsive (CPU bound > > threads get a priority reduction, foreground applications get a boost, > > etc.), but in the presence of multiple CPU bound threads attached to > > message queues (which is an application design problem), there's not a > > whole lot you can do. > > I have never seen this work properly in W2K...once you have one compute- > bound thread running, the GUI becomes totally unresponsive. Nextstep, now > that is another matter. Unless the thread is running at high priority, or is holding a lock on message processing or some common GDI item, I've never seen a CPU bound thread make the GUI "totally unresponsive" in the NT line of Windows. Slow it down, sure.
From: Jan Vorbrüggen on 8 Sep 2005 03:42 > Unless the thread is running at high priority, or is holding a lock on > message processing or some common GDI item, I've never seen a CPU bound > thread make the GUI "totally unresponsive" in the NT line of Windows. > Slow it down, sure. Well, if changing window focus takes up to ten seconds, or typing means that you get your visual feedback several seconds later all in one batch, is that slowing down or being unresponsive? There was this IBM study some decades ago about response lags in the sub-second range having a big impact on productivity... Jan
From: JJ on 8 Sep 2005 04:12 Jan Vorbrüggen wrote: > > Unless the thread is running at high priority, or is holding a lock on > > message processing or some common GDI item, I've never seen a CPU bound > > thread make the GUI "totally unresponsive" in the NT line of Windows. > > Slow it down, sure. > > Well, if changing window focus takes up to ten seconds, or typing means > that you get your visual feedback several seconds later all in one batch, > is that slowing down or being unresponsive? There was this IBM study some > decades ago about response lags in the sub-second range having a big impact > on productivity... > > Jan I see multi second lags all the time, its not just infuriating, its hair pulling ... I have 2 large monitors, an nVidia AGP and an oldish Banshee PCI card both run at or near 2k pixels. Dragging almost any large window from PCI to AGP, wait precisely 7 secs for the mess to clear up (even plain text). The obvious solutions are unpalatable. I don't think multicores will help much either. I get the distinct feeling that the increased workload being asked of W2K by the bloated office desktop SW is far beyond Moores law of so called improvement. The only thing PCs generally actually deliver on is the media delivery, VLC good example. But actually try to do more mundane stuff on larger docs, and they bog down. I suspect its got alot to do with the poor data locality and the memory wall at play against everything that doesn't look like cache friendly media codecs. BeOS, now that is another matter. johnjakson at usa ...
From: Skybuck Flying on 8 Sep 2005 10:53 "Ken Hagan" <K.Hagan(a)thermoteknix.co.uk> wrote in message news:dfjnm8$rr0$1$8300dec7(a)news.demon.co.uk... > Skybuck Flying wrote: > > > > I would want a dual core processor, I would even want a 1000 core processor. > > > > The reason is simply because then I as a programmer can start writing > > *serious* multi threaded software which can take adventage of multi core > > processors. > > A nit pick... > > You can write such software now. Your customers will need to > wait until they have such a processor before they will benefit > from it. > > Until then, there is no advantage to parallel software, only > overheads and cost. That's why we aren't all doing it already. Exactly, So chicking and egg problem. We dont write multi threaded software because it's overhead. They don't make paralleler hardware because there is no multi threaded software lol. Ofcourse that's a slighty exaggerated there is some multi threaded software out there ;) Bye, Skybuck.
From: Skybuck Flying on 8 Sep 2005 10:56
"Ken Hagan" <K.Hagan(a)thermoteknix.co.uk> wrote in message news:dfjnm8$rr0$1$8300dec7(a)news.demon.co.uk... > Skybuck Flying wrote: > > > > I would want a dual core processor, I would even want a 1000 core processor. > > > > The reason is simply because then I as a programmer can start writing > > *serious* multi threaded software which can take adventage of multi core > > processors. > > A nit pick... > > You can write such software now. Your customers will need to > wait until they have such a processor before they will benefit > from it. > > Until then, there is no advantage to parallel software, only > overheads and cost. That's why we aren't all doing it already. Anyway besides from my previous post I wanted to add a point but in my enthousiastic look up of a difficult word I pressed the post button a little bit too quickly lol... Though my point should be obvious from previous threads. The point is ofcourse that at some point in time all software will be multi threaded/parallel because it's a natural way to get more performance. Thus being a parallel processor now will be more suited to execute the software of the future =D That's gotta be worth something =D Bye, Skybuck. |