Question Regarding 09-268 SPL - STP Extended Buffer Support

Oganessyan, Gourgen goganess at intersil.com
Mon Nov 2 13:42:51 PST 2009


* From the T10 Reflector (t10 at t10.org), posted by:
* "Oganessyan, Gourgen" <goganess at intersil.com>
*
I do not have this data, I believe that's what Jim is proposing to share, for
copper cables. I understand that SAS-2 handles this by specifying maximum
prop delay for a 10m cable as 53 ns, even though the nominal number is more
like 45. Optics folks said 470 ns for a 100m cable, so again, assuming 530 ns
should give enough padding. So my proposal would be to assume 5.3 ns/m delay
irrespective of the media, which should give enough margin for any variations
and any additional delay due to circuitry. That would keep this part of it
simple, since there are architectural issues we still need to resolve. Using
5.3 ns/m instead of 4.5 ns/m only results in additional 4 dwords needed to
support a 25m cable.
But then of course other people are welcome to chime in.
Gourgen Oganessyan
Staff Product Marketing Engineer
Intersil Corp.
Phone: (408)-890-1700
	   (630)-802-0574 (cell)
Fax:	 (630)-364-5724
e-mail: goganess at intersil.com
-----Original Message-----
From: Olawsky, Barry [mailto:Barry.Olawsky at hp.com] 
Sent: Monday, November 02, 2009 3:16 PM
To: Oganessyan, Gourgen; Tim Symons; McGrath, James
Cc: t10 at t10.org
Subject: RE: Question Regarding 09-268 SPL - STP Extended Buffer Support
Gourgen,
Do you have any data on the variation of flight time with respect to the
material composition of the bulk cable? Flight time is proportional to the
reciprocal of the square root of the dielectric constant of the media. What
are the optical media options and how much does the dielectric constant vary?
Other parameters such as the equalizer circuitry delay are important too but
we can't ignore possible variations in the media.
-----Original Message-----
From: Oganessyan, Gourgen [mailto:goganess at intersil.com] 
Sent: Monday, November 02, 2009 2:55 PM
To: Tim Symons; Olawsky, Barry; McGrath, James
Cc: t10 at t10.org
Subject: RE: Question Regarding 09-268 SPL - STP Extended Buffer Support
I got some input from a couple of optical cable suppliers when we first
launched this effort, and people seemed to agree that the delay introduced by
the transmission channel is in hundreds of nanoseconds (~500ns for a 100m
cable), while the delay of the EO conversion circuitry is in sub-nanosecond
range. If that is true, I would favor just calculating the delay from the
length number (irrespectively of the transmission medium), and any added
delay should be covered by the round-up in the buffer size formula. This
would be the simplest approach. 
Also, since it appears that there is no extra room in the QSFP memory map to
call out delay, we should probably modify the buffer size formula to use the
length number.
Gourgen Oganessyan
Staff Product Marketing Engineer
Intersil Corp.
Phone: (408)-890-1700
	   (630)-802-0574 (cell)
Fax:	 (630)-364-5724
e-mail: goganess at intersil.com
-----Original Message-----
From: owner-t10 at t10.org [mailto:owner-t10 at t10.org] On Behalf Of Tim Symons
Sent: Monday, November 02, 2009 2:18 PM
To: Olawsky, Barry; McGrath, James
Cc: t10 at t10.org
Subject: RE: Question Regarding 09-268 SPL - STP Extended Buffer Support
* From the T10 Reflector (t10 at t10.org), posted by:
* "Tim Symons" <Tim_Symons at pmc-sierra.com>
*
Pd (total cable assembly prop delay) needs to report the delay from one
electrical connector the other. 
For optical cables, the bridge to laser light and back to electrical
domain should be considered part of the "active component of the cable"
and must be included in the propagation delay value.
It would be useful to have some reference data for optical cables as
well as copper. Maybe 50m, 100m, 250m and 500m optical cables Pd values
would be useful.
Tim.
Tim Symons
Principal Engineer, PMC-Sierra Ltd.
Cell: 778 998 5025
E-mail Tim_Symons at pmc-sierra.com
-----Original Message-----
From: Olawsky, Barry [mailto:Barry.Olawsky at hp.com] 
Sent: Monday, November 02, 2009 12:08 PM
To: Tim Symons; McGrath, James
Cc: t10 at t10.org
Subject: RE: Question Regarding 09-268 SPL - STP Extended Buffer Support
Tim (Symons),
I'm working with some bulk cable suppliers to calculate my own numbers.
I'm not sure how to approach the optical configurations.
-----Original Message-----
From: Tim Symons [mailto:Tim_Symons at pmc-sierra.com] 
Sent: Monday, November 02, 2009 1:55 PM
To: McGrath, James; Olawsky, Barry
Cc: t10 at t10.org
Subject: RE: Question Regarding 09-268 SPL - STP Extended Buffer Support
At an earlier meeting, it was noted that optical cables and any cables
different transmission media may exhibit propagation delays that would
be significantly different from the 5nS per meter "rule-of-thumb" used
today.
Can anyone comment on the propagation delay ranges for cables made of
other materials in addition to Jim's 24AWG/30AWG copper gauge data.
Note that this data is for managed cables Mini-SAS HD and QSFP+
Regards
	Tim.
Tim Symons
Principal Engineer, PMC-Sierra Ltd.
Cell: 778 998 5025
E-mail Tim_Symons at pmc-sierra.com
-----Original Message-----
From: McGrath, James [mailto:jmcgrath at email.cinch.com] 
Sent: Monday, November 02, 2009 10:54 AM
To: Olawsky, Barry; Tim Symons
Cc: t10 at t10.org
Subject: RE: Question Regarding 09-268 SPL - STP Extended Buffer Support
Barry, Cinch has done some work to help answer this question. For
example, in a 3m cable there is less than 0.50nanosec difference in prop
delay from a 24AWG to a 30AWG cable. So length appears to be a practical
parameter to determine prop delay. I can show a few charts, if needed,
at the meeting next week. 
Jim McGrath
Cinch Connectors
1700 Finley Road
Lombard, IL 60148
office: 630-693-2040
mobile: 630-244-3872
-----Original Message-----
From: owner-t10 at t10.org [mailto:owner-t10 at t10.org] On Behalf Of Olawsky,
Barry
Sent: Monday, November 02, 2009 11:30 AM
To: Tim Symons
Cc: t10 at t10.org
Subject: Question Regarding 09-268 SPL - STP Extended Buffer Support
* From the T10 Reflector (t10 at t10.org), posted by:
* "Olawsky, Barry" <Barry.Olawsky at hp.com>
*
Tim (Symons),
During the SFF-8636/8449 conference call on October 22nd a question was
posed for which we need your assistance to answer. Tom Palkert of
Luxtera proposed using cable length to estimate the end-device to
end-device propagation delay. To determine if this is a feasible
technique, we need to know the maximum acceptable total propagation
delay error including resolution and accuracy errors. Since you propose
limits on this parameter in 09-268 we thought you might have valuable
input on this topic.
Barry Olawsky
Hewlett Packard
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
The information contained in this electronic mail message is 
privileged and confidential information, may be subject to the 
attorney-client privilege and is intended solely for the use of 
the addressee. If you are not the intended recipient, any 
disclosure, copying, distribution, or the taking of any action 
in reliance on the contents of this message is strictly 
prohibited. If you have received this message in error, please 
notify me immediately. 
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org
*
* For T10 Reflector information, send a message with
* 'info t10' (no quotes) in the message body to majordomo at t10.org



More information about the T10 mailing list