Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!olivea!apple!agate!ucbvax!IBSUN.ABO.FI!mats
From: mats@IBSUN.ABO.FI
Newsgroups: comp.sys.transputer
Subject: (none)
Message-ID: <9101230720.AA03071@ibsun.abo.fi>
Date: 23 Jan 91 07:20:02 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 37

's message of 22 Jan 91 10:09:00 EDT <9101221515.AA11853@theory.TN.CORNELL.EDU>
Subject: Re:C004 crossbar communication delays


Ken Koontz writes:

  >> you find the following difference in throughput rate (links at 20Mbps,
  >> using fairly large buffers >= 1Kbyte):
  >> 
  >>     without C004 : 1.75 Mbytes/sec
  >>     with    C004 : 1.33 Mbytes/sec
  >>
  >> Inmos does acknowledge this fact now, but back then it was a surprise to
  >> some even at Inmos. Things only get worse if you go through 2 C004s.
  >> I haven't verified it myself but Inmos claims it goes down to around
  >> 0.8 Mbytes/sec.

We also measured the overhead caused by using C004 switches. We used
unidirectional communication on 20 Mbps links between T800 transputers
with 250 ns DRAM. The packet size was 1024 bytes. We didn't use internal
RAM. These are the results:

	  Nr. of C004's		Kbytes/s
	  -------------		--------
		0		 1715
		1		 1356
		2		 1060
		3		  871

Each C004 switch seems to decrease the communication througput with
about 20 percent.


Mats Aspnas
Abo Akademi,
Dept. of Computer Science
email: mats@ibsun.abo.fi