xoomsys (just trying my luck)

[fogh]

Hi All,

Anyone who can share some info on xoomsys ?

--
Frederic
oghdayan at x.s.4.a.l.l dot nl

 

[tstengerster@gmail.com]

I called them a few weeks ago to check out their product. They are not
a simulator, they are a wrapper to your current simulator. It only
works with Time domain simulators right now. Tool made for extracted
PLL circuits and extracted digital clock circuits.

If using spectre, for example, - the desiger really does not see this
tool. He would netlist out of ADE, produce a netlist. Then this tool
will analize the netlist, decide how to "break it up" and spawn off the
parallel simulation jobs (tries to balance the load across all CPUs),
after all the sims complete, the tool reassembles the data, and the
user would plot the waveforms in wavescan (or waveform viewer of
choice).

Hope that helps,

Tamara
I have not seen it in action. It sounded like they were a month to 2
months away from when we could get a tool in house to look at. They
are currently working with a few beta sites. They did offer to take
one of our circuits and run it for us.

 

[fogh]

tstengerster@gmail.com wrote:

I called them a few weeks ago to check out their product. They are not
a simulator, they are a wrapper to your current simulator. It only
works with Time domain simulators right now. Tool made for extracted
PLL circuits and extracted digital clock circuits.

If using spectre, for example, - the desiger really does not see this
tool. He would netlist out of ADE, produce a netlist. Then this tool
will analize the netlist, decide how to "break it up" and spawn off the
parallel simulation jobs (tries to balance the load across all CPUs),
after all the sims complete, the tool reassembles the data, and the
user would plot the waveforms in wavescan (or waveform viewer of
choice).

Hope that helps,

Tamara
I have not seen it in action. It sounded like they were a month to 2
months away from when we could get a tool in house to look at. They
are currently working with a few beta sites. They did offer to take
one of our circuits and run it for us.

Tamara,

thanks. What you write here confirms the little that can be found in internet.

If the partitioning creates completely independent spectre jobs, I do not
understand how it can partioning anything except completely disconnected graphs
(such as a schematic where the same testbench is copy/pasted all over the page
to make only slight alterations.) I can t fathom for instance how this can
partition a PLL.

Regards,
--
Frederic