<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=us-ascii">
<META content="MSHTML 6.00.2900.2802" name=GENERATOR></HEAD>
<BODY
style="WORD-WRAP: break-word; khtml-nbsp-mode: space; khtml-line-break: after-white-space">
<DIV dir=ltr align=left><SPAN class=180355320-09032006><FONT face=Arial
color=#0000ff size=2>Infiniband with DDR is already at 20Gbps over CX4
copper</FONT></SPAN></DIV><BR>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
<HR tabIndex=-1>
<FONT face=Tahoma size=2><B>From:</B> beowulf-bounces@beowulf.org
[mailto:beowulf-bounces@beowulf.org] <B>On Behalf Of </B>Ed
Karns<BR><B>Sent:</B> Thursday, March 09, 2006 10:35 AM<BR><B>To:</B>
beowulf@beowulf.org<BR><B>Subject:</B> [Beowulf] Re: newbies' dilemma /
firewire? (Hahn)<BR></FONT><BR></DIV>
<DIV></DIV><BR>
<DIV>
<DIV>On Mar 8, 2006, at 12:00 PM, <A
href="mailto:beowulf-request@beowulf.org">beowulf-request@beowulf.org</A>
wrote:</DIV><BR class=Apple-interchange-newline>
<BLOCKQUOTE type="cite">
<P style="MARGIN: 0px"><FONT style="FONT: 13px Lucida Grande"
face="Lucida Grande" size=4>... that's nice marketing-speak, but content-free,
I'm afraid. ... though I have to say I really don't expect
optical </FONT><FONT style="FONT: 13px Lucida Grande"
face="Lucida Grande" size=4>WDM firewire to make any difference ;)
... I'd personally love to see something better in every way than
GBE/10GE, </FONT>but just don't see it happening. no, please, don't
suggest IB :)</P></BLOCKQUOTE></DIV><BR>
<DIV>
<DIV>Well, it can, does and will make a difference "real soon, now", considering
that since metal conductors are close to the upper throughput limits dictated by
the laws of physics and laws of diminishing returns. Pushing
more data through a metal conductor, beyond several Gigabits per second,
requires more and more power over shorter and shorter metal conductors ... and
considering power (heat) and cross talk (noise) v. conductor length (of antenna)
... you sure don't want to open an active box as the microwave radiation could
fry your gonads (resulting legal action being the main reason no one is
seriously promoting 20 or 30 Gigabit Ethernet over wires).</DIV>
<DIV><BR class=khtml-block-placeholder></DIV>
<DIV>Intel (et al) is seeking alternates to the above: <A
href="http://www.intel.com/technology/silicon/sp/">http://www.intel.com/technology/silicon/sp/</A>
... data transfer over packet switching networks using multiple frequencies
(multiplexed "colors") and approaching Terabits per second over a silicon glass
fiber bus.</DIV>
<DIV><BR class=khtml-block-placeholder></DIV>
<DIV>Key point: determination of the protocol of choice for this "Glass Bus".
Candidates include packet switching and non-packet switching protocols:
FireWire, USB, SATA, PCI (e or other), Fibre Channel, iSCSI / SAS, <FONT
class=Apple-style-span face=Verdana>InfiniBand, </FONT>EtherNet and
possibly others ... and considering that FireWire is the most hardware/ firmware
efficient for a given microprocessor speed, having the most hardware efficient,
packet switching network and being peer to peer topography ... </DIV>
<DIV><BR class=khtml-block-placeholder></DIV>
<DIV>Thus the previous discussion(s) Re: " ... Why is it comparable to or
faster from a processor running at around 40% of GigaBit processor speeds?
Processor efficiency FireWire has a 32-bit "risc" type microprocessor, is
peer to peer in hardware / firmware and has other lean architecture
features. (Small address space, data frame large = more efficient data
packet over double duplex connections.)</DIV>
<DIV><BR class=khtml-block-placeholder></DIV>
<DIV>(Picture this: a coffee can containing multiple processors on multiple
circuit boards in close proximity fanning out from a central glass rod or fiber
running through the axis ... linking to other coffee cans ...)</DIV>
<DIV><BR class=khtml-block-placeholder></DIV>
<DIV>EOR = end of rant</DIV>
<DIV><BR class=khtml-block-placeholder></DIV>
<DIV>Ed Karns</DIV>
<DIV>FireWireStuff.com </DIV><BR
class=Apple-interchange-newline></DIV></BODY></HTML>