Re: [IBIS-Users] IBIS - AMI - questions

Anad,

An IBIS-AMI simulation is performed in two pstages:

1) In the first stage, the _analog network_ is characterized to derive 
an impulse response.  The analog network is defined as the _passive 
network_ (everything between the pads) *and* the analog characteristics 
of the transmitter's output (voltage, slew rate, output impedance, 
capacitance) along with the receiver's input termination network.  The 
traditional .ibs file can be used for this analysis.  This is called the 
_analog portion_ of the IBIS-AMI model.  In truth, the existing .ibs 
syntax falls a bit short of providing good broadband characterization of 
the TX output and RX termination network, which is why SiSoft and IBM 
are proposing enhancements to the standard ... but you get the idea.

2) Once the impulse response for the analog network has been derived, 
the second part of IBIS-AMI simulation involves Channel Analysis, which 
uses a new class of convolution-based simulation algorithms.  The 
_algorithmic part_ of the IBIS-AMI model is supplied as a binary object 
that is linked into the Channel Simulator and called at runtime.  the 
IBIS-AMI standard covers what waveform types can be passed to / returned 
by the Algorithmic Model, the details of the binary calling interface 
and how control parameters are passed to the model.

There are a number of different analysis flows possible with IBIS-AMI 
Algorithmic Models - more than I can get in to here.  High points:

Statistical Analysis: IBIS-AMI models can be used to perform a 
linearized convolution analysis similar to StatEye, with models from 
semiconductor vendors that represent actual IP.  Statistical Analysis is 
extremely fast, and hundreds of simulations can be run in a few minutes.

Time-Domain Analysis - supports modeling of non-linear  equalization (RX 
DFE) and time-varying behavior  (automatic tap adaptation algorithms, 
clock recovery algorithms).  Simulation performance is on the order of 
1,000,000 bits/ minute.

The free SiSoft IBIS-AMI toolkit includes the executable .dll, the .ibs 
file and the .ami (algorithmic model control) file.  The standalone 
simulator executes a single algorithmic model in a test bench 
environment.  The standalone simulator does not use the analog model in 
the .ibs file to derive the impulse response - you have to provide that 
through other means.

Anand, I hope this provides an overview that helps put things in 
context.  We're happy to discuss in more detail if you want.

Todd.

Todd Westerhoff
VP, Software Products
SiSoft
6 Clock Tower Place, Suite 250
Maynard, MA 01754
(978) 461-0449 x24
twesterh@sisoft.com
www.sisoft.com



Muranyi, Arpad wrote:
> Anand,
>
> The only examples I am aware of on the web are the
> SiSoft and Cadence AMI evaluation tool kits.
>
> These have a Tx and an Rx model, including their
> source code, and an executable which will let you
> run them.  There is no IBIS file there, since as
> I told you before, there is really not much in an
> IBIS file other than a pointer.
>
> http://www.vhdl.org/pub/ibis/macromodel_wip/archive/20080523/sisoft/SiSo
> ft_IBIS-AMI_Eval_Toolkit_v2_10.zip
> http://www.vhdl.org/pub/ibis/macromodel_wip/archive/20080523/cadence/Cad
> ence_IBIS_AMI_Evaluation_Toolkit_v2_1a.zip
>
> Unfortunately these examples do not have a "healthy"
> example of a parameter file (.ami) so you can't
> use them to learn proper .ami file syntax.  The
> SiSoft example uses a "proprietary" CSV file format
> for entering the parameters and the Cadence example
> uses an unofficial subset of the .ami tree format.
>
> Arpad
> ==================================================== 
>
> -----Original Message-----
> From: Anand Ramalingam [mailto:anandr@Magma-DA.COM] 
> Sent: Friday, June 12, 2009 3:48 PM
> To: Muranyi, Arpad; ibis-users@eda.org
> Subject: RE: [IBIS-Users] IBIS - AMI - questions
>
> Hi Arpad,
>
> Thanks for your prompt reply and the pointers. 
>
> [Arpad]
>
> 3)  A full fledged .IBS file alone is not going to help you much
>     because there is really not much in the .IBS file other than a
>     pointer to the AMI DLL and its parameter file (.ami).  What you
>     need is a good example (i.e. a working DLL) with an appropriate
>     parameter file (.ami) and a simulator that can run them.
>
> [Anand] By any chance do we have a complete example (with ibis/dll/ami
> files) on the web which I can download and simulate?
>
> Thanks
> Anand
>
>   

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

--------------------------------------------------------------------
|For help or to subscribe/unsubscribe, e-mail majordomo@eda-stds.org
|with the appropriate command message(s) in the body:
|
|  help
|  subscribe   ibis       <optional e-mail address, if different>
|  subscribe   ibis-users <optional e-mail address, if different>
|  unsubscribe ibis       <optional e-mail address, if different>
|  unsubscribe ibis-users <optional e-mail address, if different>
|
|or e-mail a request to ibis-request@eda-stds.org.
|
|IBIS reflector archives exist under:
|
|  http://www.eda-stds.org/pub/ibis/email_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/users_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/email/         E-mail since 1993