- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello all,
I am curious about the following:
Based on the latency and throughput numbers for the Pentium 4, I don't see an obvious advantage of fixed point computations over single precision floating point (specifically for multiplication and division operations).
Is this a valid statement?
Is there any speed benefit in converting floating point code to fixed point?
Thanks much,
-Mahesh
I am curious about the following:
Based on the latency and throughput numbers for the Pentium 4, I don't see an obvious advantage of fixed point computations over single precision floating point (specifically for multiplication and division operations).
Is this a valid statement?
Is there any speed benefit in converting floating point code to fixed point?
Thanks much,
-Mahesh
Link Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Specifically for multiplication and division operations, fixed point will have a serious performance disadvantage on P4. Unless you are supporting a CPU which lacks hardware floating point instructions, you are unlikely to see a benefit in fixed point.
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