- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
What is theoverhead in HT processors compared to traditional processors for
context switching,at least using Win32 and C++?
It will be better for students like me...because academia is not interested in MS technologies... Anything in C++ will be much more interesting than C#... :D
Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I don't think a useful answer is possible without more definition of what you mean. The HT related issue of giving one or the other of 2 running threads priority on shared resources, compared with 2 threads running on a single core CPU, is of most interest when supporting a few specific applications such as Oracle. It was more interesting when there were more HT than dual core CPUs. On the dual core CPUs, there may be shared cache and bus resources, but the term context switching has even less meaning in the competition for those resources.
If you are talking about the use of the HT terminology for Itanium switch on event multi-tasking, the overhead for switching there is significantly greater than for past Xeon HT CPUs.
If you are talking about traditional meaning of context switching, it's hard to see how HT is related, unless you have a specific situation in mind.
If you are talking about the use of the HT terminology for Itanium switch on event multi-tasking, the overhead for switching there is significantly greater than for past Xeon HT CPUs.
If you are talking about traditional meaning of context switching, it's hard to see how HT is related, unless you have a specific situation in mind.

Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page