tulip vs. Lite On

[Douglas Eadline]

On Tue, 8 Dec 1998, Robert G. Brown wrote:

> On Tue, 8 Dec 1998, Douglas Eadline wrote:
-snip-- the data

> These are very interesting numbers, Doug.  Have you tried tulip->lite-on
> or lite-on->tulip?  They are interesting because the card is handling
> only ~6000 pps on the transmission side!  This is absurdly low (0.17 ms
> between packets) since the transmissions ought not to be blocked by wire
> bandwidth.

lite-on to tulip:
==================

UDP UNIDIRECTIONAL SEND TEST to matthes8
Socket  Message  Elapsed      Messages                
Size    Size     Time         Okay Errors   Throughput
bytes   bytes    secs            #      #   10^6bits/sec

 65535    1472   60.00      363875      0      71.42
 65535           60.00      363875             71.42

TCP STREAM TEST to matthes8
Recv   Send    Send                          
Socket Socket  Message  Elapsed              
Size   Size    Size     Time     Throughput  
bytes  bytes   bytes    secs.    10^6bits/sec  

 65535  65535  65535    60.00      71.06   

tulip to lite-on:
=================

UDP UNIDIRECTIONAL SEND TEST to matthes2
Socket  Message  Elapsed      Messages                
Size    Size     Time         Okay Errors   Throughput
bytes   bytes    secs            #      #   10^6bits/sec

 65535    1472   59.99      487411      0      95.68
 65535           59.99      487411             95.68

TCP STREAM TEST to matthes2
Recv   Send    Send                          
Socket Socket  Message  Elapsed              
Size   Size    Size     Time     Throughput  
bytes  bytes   bytes    secs.    10^6bits/sec  

 65535  65535  65535    60.00      94.76   

> 
> Another question:  have you tried a couple or three points along the
> small-message part of the curve?  You might try (UDP_STREAMS) m = 1, m =
> 512, m = 1024 to sort of sketch out how the NIC handles small packets --
> on a real tulip the interface should handle 20K-50K pps for small
> packets (depending on host CPU speed) until the interface starts to
> saturate the available bandwidth and has to cut back on the number of
> packets sent.  One is curious -- if the lite-on can only send 6K pps
> with optimal packet sizes (which is far from saturating the bandwidth),
> can it only send 6K pps for small packets?  

I do not have a whole lot of time today (like when do I have time).
Actaully, I was going to try netpipe which gives a graph of the
performance.

> 
> If so, this would make the NIC absurdly poor, far worse than only "75%
> of a tulip".  In fact, I would call it openly "broken" unless there is a
> lite-on-specific card hack unknown (so far) to Don that might make it
> perform normally.  Alas, all these tulip-chip clones and even many real
> tulip cards layer on the nonsense that makes it difficult to impossible
> to write a "generic" tulip driver; Don has expressed his frustration on
> this very subject but what can you do?
> 
Well, I'm sure it could be a driver problem.  I just tried "out of the
box" measurements because the assumtion was the lite-on was plug and
play.

Doug


-------------------------------------------------------------------
Paralogic, Inc.           |     PEAK     |      Voice:+610.861.6960
115 Research Drive        |   PARALLEL   |        Fax:+610.861.8247
Bethlehem, PA 18017 USA   |  PERFORMANCE |    http://www.plogic.com
-------------------------------------------------------------------