( ESNUG 477 Item 7 ) -------------------------------------------- [11/20/08]
Subject: ( ESNUG 476 #1 ) User questions if Sierra MCMM is tipping point
> We use the 3 major P&R tools (ICC, SOCEnc, Talus) here, but personally I
> have used only Talus and SOCEncounter. I would definitely consider Sierra
> if I had to do a fresh P&R tool evaluation.
>
> - Nanda Lekkelapudi
> MIPS Technologies, Inc. Mountain View, CA
From: [ The Invisible Man ]
Hi, John,
Our group uses Cadence Encounter. Through acquisitions we now have groups
that also use Synopsys Astro and Magma Blast. There is no consensus about
which is better. Each group tends to favor the tool they know.
None of the older tools were fundamentally architected for MCMM. The
question is how important is that? How difficult is it to swap in a new
timing engine? I don't really know the answer. Mentor (and AtopTech) claim
it is difficult. My instinct tells me it isn't trivial since nowadays
timing reaches into every corner of the P&R flow. The current tools mostly
came into being when the replaced earlier non-timing-driven tools. Mentor
argues, somewhat pervasively, that that there is a new discontinuity related
to problems that are showing up at the smaller geometries.
SNPS and LAVA are each going through an interesting transition: SNPS to ICC
and LAVA to Talus. Their decisions to go to a next gen architecture would
seem to support the argument that their own tools needed some new data
structures underneath to support the requirements of 65 nm and beyond.
Switching tools is very difficult. In addition to the cost of the tools
themselves there is the cost in training and productivity. These are not
simple tools and over the years people build expertise and infrastructure
(scripts, Makefiles, etc) around a particular tool. Realistically it is
likely that we would incorporate a new tool gradually instead of a massive
fork-lift upgrade. In addition, we tend to let each group choose there own
tools instead of mandating what they use. Different groups would do
different things. I think having multiple tools also keeps the vendors
honest both technically and on the business side.
I think there is a lot of parity in the P&R space. I don't think there is
clearly a tool that is "best" or "worst". There are also a lot of metrics:
timing closure, area (utilization), power and capacity to name a few. When
you combine different designs with different tools and different expertise
of the person using the tool results can vary dramatically. The only way
to know for sure is spend time with the different tools.
Anyway, I personally am not focusing on P&R and the moment, but if I were,
Sierra Olympus would be near the top of my list of things to look at.
- [ The Invisible Man ]
Index
Next->Item
|
|
