FPGA Central

Syed,

That's a good question - it's probably worth restating what the IBIS-AMI toolkit is intended to do. The "new" part of IBIS-AMI was the introduction of the executable algorithmic model and its associated .ami control file - the analog part of the model uses the pre-existing analog syntax.

Given that we had never standardized executable models before, there was concern about whether EDA and semiconductor vendors would be able to implement the simulator/model API properly. Debugging was a potential problem - if a .dll didn't work in a given simulator, how was one supposed to know whether it was the simulator or the model that caused the issue?

The IBIS-AMI toolkit provides two components that can be used to establish whether a model or simulator conforms to the API, thus helping diagnose a compatibility problem.

The simulator test bench (IBIS_AMI_test.exe) is a standalone, single model simulator that can load and run an IBIS-AMI algorithmic model. It's used to prove that a model conforms to the API and can successfully receive and process data. It also provides a useful environment for developing, running and debugging a model standalone, which a number of people have done.

The IBIS-AMI model provides a reference model (.ibs, .ami, .dll) that a channel simulator vendor can use to test whether their simulator conforms to the API. As I mentioned this morning, customers can also use this model to determine how readily a given simulator supports the standard "out of the box."

.... which brings is back to the toolkit. As the primary purpose was to establish whether an algorithmic model's API conforms to the standard, the focus is the standalone execution of an algorithmic model. The documentation that comes with the kit says that you have to supply an impulse response (derived by whatever means) to the model as input, along with the model's configuration data. We supplied a reference impulse response in the kit for those who didn't want to come up with their own.

I believe the way we generated the impulse response is the kit was by simulating a step response for the TX / channel / RX in HSPICE. The analog model we used for the RX represented the termination behavior only. We then took the step response and differentiated it to get the impulse response we included in the kit.

Since the impulse response included in the toolkit includes the analog TX / channel / analog RX, running that impulse response with the TX algorithmic model provides the signal at the RX pad. Processing *that* waveform with a RX algorithmic model (not included in the kit) would yield the data signal at the RX sampling point.

.... that's a very long way of saying we didn't include the RX analog model in the kit because we expected people would either use the impulse response we provided, or use their own channel model / RX analog model / circuit simulator to generate their own impulse response.

Todd.


--

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


----- Original Message -----
From: "Syed Huq (huqs)" <huqs (AT) cisco (DOT) com>
To: twesterh (AT) sisoft (DOT) com
Cc: si-list (AT) freelists (DOT) org
Sent: Monday, February 8, 2010 4:27:36 PM
Subject: RE: [SI-LIST] Correction: Re: IBIS-AMI Vendor Support Help

Hi Todd,
Pls correct me if I am wrong. I believe a RX Analog front end is needed
(Along with a TX Analog IO model) for the Impulse response of the
Channel. So how is that possible in the tool kit without a RX model.

Tks
Syed

-----Original Message-----
From: Todd Westerhoff [mailto:twesterh (AT) sisoft (DOT) com]
Sent: Monday, February 08, 2010 12:13 PM
To: Syed Huq (huqs)
Cc: si-list (AT) freelists (DOT) org
Subject: Re: [SI-LIST] Correction: Re: IBIS-AMI Vendor Support Help

Syed,

The base receiver model SiSoft uses includes a peaking filter, Decision
Feedback Equalizer and full clock recovery loop. This represents a
substantial development effort and isn't something that we're able to
publish. Having a simpler receiver model that we can publish is
something that we can consider that for a future addition to the
toolkit.

I understood the task at hand to be providing a method to determine
whether a given simulator could utilize a standard IBIS-AMI model (.ibs,

..ami, .dll). I believe the reference model in the current toolkit can
be used to accomplish that task.

Thanks,

Todd.


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



Syed Huq (huqs) wrote:
> Todd,
> The referenced link only has a sisoft_tx.ibs file. Where is the
Receiver
> model ?
>
> Tks
> Syed
>
> -----Original Message-----
> From: si-list-bounce (AT) freelists (DOT) org
[mailto:si-list-bounce (AT) freelists (DOT) org]
> On Behalf Of Todd Westerhoff
> Sent: Monday, February 08, 2010 11:16 AM
> To: si-list (AT) freelists (DOT) org
> Subject: [SI-LIST] Correction: Re: IBIS-AMI Vendor Support Help
>
> All,
>
> The URL I posted this morning for the IBIS-AMI simulation toolkit and
> IBIS-AMI reference models was incorrect. The correct link is:
>
> http://www.sisoft.com/elearning_ibis-ami.asp
>
> The links to the IBIS-AMI Toolkit and reference model can be found at
> the bottom of the page. I apologize for any confusion.
>
> Todd.
>
> Todd Westerhoff
> VP, Software Products
> SiSoft
> 6 Clock Tower Place, Suite 250
> Maynard, MA 01754
> (978) 461-0449 x24
> twesterh (AT) sisoft (DOT) com
> www.sisoft.com
>
>
>
> ------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request (AT) freelists (DOT) org with 'unsubscribe' in the Subject field
>
> or to administer your membership from a web page, go to:
> http://www.freelists.org/webpage/si-list
>
> For help:
> si-list-request (AT) freelists (DOT) org with 'help' in the Subject field
>
>
> List technical documents are available at:
> http://www.si-list.net
>
> List archives are viewable at:
> http://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 (AT) freelists (DOT) org with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request (AT) freelists (DOT) org with 'help' in the Subject field


List technical documents are available at:
http://www.si-list.net

List archives are viewable at:
http://www.freelists.org/archives/si-list

Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu