- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hello all,
Was wondering if anyone has characterized the performance of the interniche IP stack throughout it's various configurations (full tcp/ip udp/ip, only upd etc) and been able to show performance gains in tuning... I am considering hw acceleration and am wondering if I will be able to tweak the system for performance (running straight udp - as stripped down as possible, with full 65,507 byte data fields) in order to avoid my next step. I'm running a gigabit link thru tse - wondering what kind of MBpS I can expect at best case... also any input on the parameters to change would be. Thanks - ChrisLink Copied
1 Reply
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
There is an Altera app note that does list various speeds over different RAM and hardware checksumming methods. We chose to use lwIP because it's faster (I don't have exact numbers but I know it to be faster) and because lwIP doesn't require uC/OS-II. Whereas Interniche "superloop" mode is hard to get running and barely documented, lwIP has an efficient "RAW API" mode that does support a superloop type of mode. lwIP also has a netconn API as well as a BSD mostly-compatible socket API.
Bill
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