( ESNUG 586 Item 3 ) ---------------------------------------------- [04/18/19]
Subject: Ravi rebukes the poor Siemens/MENT formal showing in "Best of 2018"
RAVI DROPPED THE BALL: Yes, Ravi Subramanian may have kicked some serious
ass this year with his Solido machine learning taking #1 Best of 2018 with
the users -- but the Ravi Subramanian, who's the VP & GM of All Digital IC
Validation at Siemens/MENT seriously messed up with PSS and InFact.
PSS Tool Word Counts in DeepChip "Best of" 2017 vs. 2018 EDA User Survey:
Cadence Perspec 2017: ############################### (3,097 words)
Cadence Perspec 2018: ##################################### (3,712 words)
Mentor InFact 2017: . (24 words) <-- damn it!!!
Mentor InFact 2018: #### (437 words) <-- double damn it!!!
Perspec & Breker sucked up all the user mindshare air in the PSS Wars;
and those 3 users who talk about InFact are praising InFact for being a
POINT TOOL -- plus they openly doubt InFact being PSS at all! D'oh!
- "Siemens/MENT InFact fails in the PSS Wars is Best of 2018"
http://www.deepchip.com/items/dac18-02c.html
From: [Ravi Subramanian of Siemens/Mentor]
Hi, John,
Now that my charter at Siemens/Mentor has expanded to include digital and
analog verification, I've been digesting 1's and 0's for breakfast (in
addition to my VGAs and delta-sigma modulators). I have also been reading
your "Best of 2018" series items with even greater interest.
I'm still coming up to speed on the formal verification domain, but right
off the bat I spotted some clear discrepancies in the February 14 article
vs. my view of the market, and our Questa Formal (formerly "0-In") product
line's capabilities.
The first thing that jumped out at me was the "voodoo" market share numbers
and growth trends. John, where do you get this stuff??? In short, while I
acknowledge Cadence/Jasper currently has the lion's share of the formal
market today, suffice it to say the CAGR and share numbers for Mentor formal
tools are significantly better than reported. (Our parent company does not
break out Mentor's financial results in its public financial reporting, so
I cannot share more details.)
While Spy #2 in 2016 said the $84 million 2016 ABV market broke out to:
Cadence Jasper : :######################### $50M (60%)
Mentor Questa FV : :########## $17M (20%)
OneSpin : :##### $11M (13%)
Synopsys VC Formal : :### $6M (7%)
But now 2 years later in 2018, using these ABV user comments as a proxy
for market dollars, the 2018 ~$100M histogram now ballparks out to:
Cadence Jasper : :################################## $67M (67%)
Synopsys VC Formal : :######## $17M (17%)
OneSpin : :###### $10M (10%)
Mentor Questa FV : :## $4M (4%)
However, I can say that in your report, your Spy #2's market share numbers
for 2016 which had us ~$17 M with 17 percent share were very close to actual
numbers back in 2016. While I'm prohibited from sharing the real numbers
for 2017 and 2018, I can confirm that we have seen nice revenue growth over
the last two years -- and certainly not a decline -- so stating our revenue
went from $17 M to $4 M for the MENT formal product line is completely
inaccurate. I'd call it fake news, but I'm too polite to say that.
Conclusions: ...
2. Synopsys VC Formal has moved from a distant #4 in 2016, to now
in 2018 being the #2 in ABV.
3. OneSpin and Questa FV are in the much weaker #3 and #4 slots.
And conclusions 2 and 3 are also inaccurate.
On a different note, based on data from the ESDA market reports as well as
our own independent annual survey of over 2,000 EDA users conducted by the
Wilson Group, I do agree that the formal verification market growth is
outpacing the overall EDA market; and the estimated TAM of $85 - $100M is
in-line with our internal estimates.
On the technical front (aka John's tool feature smackdown), out of the 10
technical features cited in the article, my R&D people confirm we have had
6 of them since "0-In days", and 2 others for several years now.
So it is incorrect to claim that these capabilities are unique to Jasper,
or are even new to the industry. And yes, John, I concede there are some
points where we still have some gaps to close.
---- ---- ---- ---- ---- ---- ----
MACHINE LEARNING AT SIEMENS/MENTOR
Ravi's newly aquired Solido machine learning was 100% at the right place
and right time as it applies to EDA tools -- plus Ravi and Amit knew how
to spread ML through a massive Siemens customer base -- thus earning
it an exceedingly rare user-given MERGERS THAT WORKED Awards -- and
also gaining the DeepChip User-Voted #1 Best of 2018 Award, too! Wow.
As you know, Machine Learning (ML) is near-and-dear to my heart - it's the
incredibly effective secret sauce inside our Solido variation-aware design
and characterization flows.
Naturally, Mentor as a whole is looking to leverage this technology across
all of our analog and digital flows to deliver verification insight and
enhance verification efficiency. Stay tuned for some cool stuff on that!
Regarding formal and ML specifically: some of the algorithms and techniques
that have been in our Questa PropCheck tool for years -- like its data-based
engine-tuning -- is actually a form of ML technology.
Indeed, our Questa PropCheck R&D team is constantly investigating areas
where various ML techniques could be game changers. What I can say is
that in reviewing the non-ML vs. the ML-based approaches, "the math" right
now favors our current engine orchestration architecture compared to rival
competing approaches. But we have no religion: if any new techniques start
to outperform our current formal engines, we will happily adopt them.
---- ---- ---- ---- ---- ---- ----
To summarize:
* You really need better spies -- we have a much higher share of this
growing market than what you're seeing.
* Our "0-In" team and product line is highly competitive and thriving
under the Questa Formal banner -- and not dead like you suggest.
* It's clear that Machine Learning has tremendous potential to improve
many facets of verification, but it's still early days. And we'll
add it when it measurably improves verification -- not just to add
it as a marketing buzz word.
I look forward to seeing you at this upcoming DAC'19 in Las Vegas.
- Ravi Subramanian
Mentor/Siemens Fremont, CA
---- ---- ---- ---- ---- ---- ----
Related Articles:
Siemens/MENT InFact fails in the PSS Wars is #2c for Best of 2018
Mentor InFact was pretty much a No Show in the 2017 PSS tool wars
CDNS Perspec is crushing it in the PSS Wars is #2 for Best of 2018
User buzz on Siemens/Solido machine learning is #1 for Best of 2018
Join
Index
Next->Item
|
