[SI-LIST] Re: source of DDR 4 timing recommendations
- From: "Chauhan, Rajat" <rajat.chauhan@xxxxxx>
- To: "Hermann.Ruckerbauer@xxxxxxxxxxxxx" <Hermann.Ruckerbauer@xxxxxxxxxxxxx>, "ericsilist@xxxxxxxxx" <ericsilist@xxxxxxxxx>, "si-list@xxxxxxxxxxxxx" <si-list@xxxxxxxxxxxxx>
- Date: Sat, 2 Aug 2014 18:44:59 +0000
At very high data rates, like DDR4, controllers comes with per DQ bit deskew
feature. Each bit is internally adjusted to cancel out the static skew (skew
across the link, slicon+package+pcb) between DQs.
The 90ps may be the max skew controller can adjust. You may need to check with
your vendor.
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx] On
Behalf Of Hermann Ruckerbauer
Sent: Saturday, August 02, 2014 9:09 PM
To: ericsilist@xxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: source of DDR 4 timing recommendations
Hello Eric,
most of the design guides on DDRx memory use very strict requirements on the
lenght matching.
This comes from the fact that many people think this is an easy task and comes
for free. But I have seen several designs that have been length matched to
death.
Not sure where the 50mil and 90ps come from .. 1.25mm (50mil) are below 10ps so
quite small, while 90ps is quite a bit for the given UI ..
I have seen the 1mm in several design guides and usually relax these to 5mm
(~35 ps) .. at least for DDR3 this is usually acceptable, while I would not
give away the 90ps that you mentioned also ..
So you should make your own thoghts and check what is the real requirement for
your system.
But of course this means you have to understand the signaling to make the right
judgement.
DDR4 changes the concept a bit to DDR3 as it tries to work with BER with some
RX mask and tdiVW/VdiVW, what was not considered in DDR3.
And you will find numbers for the input valid window also in the JEDEC spec
just approaching this from a conventional point of view the high level
distribution of uncertainties is 1/3 for the TX (e. g. output uncertainty for
the controller), 1/3 fo the RX (e. g. DRAM setup/hold) and 1/3 for the channel
(e. g. ISI and X-talk effects). Building up a timing budget will help you to
understand what is allowed in your simulation. ..
If some numbers are not in the JEDEC spec you should take your vendors
datasheet as source (I always look at both when working with DRAM Memory). Some
vendors will give quite a bit more data and explanation on Valid window RX Mask
and setup/hold times.
Regarding a good explanation on DRAM input timings: I think this is difficult
via e-mail and the reason why I do the DRAM seminar. So I think you have to do
a very thorough reading of the JEDEC and your vendors spec in order to build up
the understanding of the DDRx signaling ..
Hermann
Upcoming Events:
=================
"Open the Black Box of Memory"
What you always wanted to know about Memory!
.. But never had the right expert to ask!
September 24./25. 2015 in Copenhagen
vist www.EyeKnowHow.de/en/seminars/
EKH - EyeKnowHow
Hermann Ruckerbauer
www.EyeKnowHow.de
Hermann.Ruckerbauer@xxxxxxxxxxxxx
Itzlinger Strasse 21a
94469 Deggendorf
Tel.: +49 (0)991 / 29 69 29 05
Mobile: +49 (0)176 / 787 787 77
Fax: +49 (0)3212 / 121 9008
Am 01.08.2014 21:49, schrieb eric silist:
> Can anyone help me understand the source of the pcb routing recommendations
> I see for DDR4. For instance DQ to DQS timing for a byte lane? I see
> some papers quoting 90ps or 50mil, but I don't know where those
> numbers come from.
> At a basic level I understand that DQS is the clock for the data byte
> on DQ, And that the total timing budget must be spread across chip, package,
> board, and dram. If I look at say DDR4-2800 that's a clock of 1.4Ghz with
> a period of 714ps. So at double data rate I figure each byte is 357ps
> wide. Then that should mean I have +/- 178ps of total margin that the
> DQS can move from it's ideal position of being in the center of the. I think
> :) Maybe that's where the 90ps for the board and 90ps for chip come from.
>
> Anyway I'm just having trouble finding a good explanation to help me
> understand how all that timing margin is calculated for each piece in
> the chain.
>
> I did try to look up DQ-DQS timing in the JEDEC standard and it just says
> TBD. I've found several docs that give me an idea of what values to use,
> but none so far that explain how they are calculated. I feel like this is
> still an SI related question since I'll use this information to setup
> constraints for my simulations but sorry if it's off topic.
>
> Thank you.
>
>
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>
>
> List forum is accessible at:
> http://tech.groups.yahoo.com/group/si-list
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
>
> Old (prior to June 6, 2001) list archives are viewable at:
> http://www.qsl.net/wb6tpu
>
>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
