( ESNUG 589 Item 06 ) --------------------------------------------- [04/22/21]
Subject: Angry Siemens 0-in R&D guy rebukes Cooley's OneSpin/0-in analysis
Will OneSpin be in charge of reinvigorating the MENT Formal ABV tools? Or
will OneSpin be subservient to MENT's 0-in olde guard and left to flounder?
Will OneSpin be allowed to keep operating as a hungry aggressive start-up
(like they have for years) that just happens to be inside Siemens EDA?
OneSpin's "going where Cadence isn't" approach is Best of 2020 #4a
OneSpin extensible ABV plus formal linting is the Best of 2020 #4c
OneSpin DV-Verify's surprise comeback in ABV is Best of 2019 #5a
OneSpin in formal areas where Cadence is NOT is Best of 2019 #5c
JasperGold and OneSpin both get #11 in 2017 for "Best of" EDA tools
OneSpin's #2 in #1 "Best of 2016", the "we try harder" company
And how JasperGold compares to OneSpin in the Formal Apps biz
Raik & Sawicki on OneSpin, Questa FV, Jasper, and formal apps
Raik on OneSpin's four formal technologies the others don't have
Raik and Hogan on formal apps, formal engines, and SNPS VC Formal
Raik on it's only OneSpin vs. Jasper in formal; not SNPS nor MENT
OneSpin CEO cites 8 "insufficiencies" in Jim Hogan's Formal Guide
Hogan on Cadence, Mentor, OneSpin, Real Intent, Synopsys formal
Or will OneSpin be Borg-absorbed into Siemens MENT corporate marketing and
left to die a slow death -- just like how 0-in (aka "Questa Formal") was?
How these questions are answered will make the difference between Anirudh
breathing a quick sigh of relief at this acquisition news -- or Anirudh
having serious headaches about ABV competition over the years.
- How CDNS COVID-19 indecisiveness gave Siemens #2 in Formal ABV
From: [ Angry Siemens 0-in R&D guy ]
Hi John,
I'm writing you from a gmail account. My work email is [xxx@xxx.com]
and my private cell number is XXX-XXX-XXXX. Feel free to call me to
confirm that I'm real, but don't copy ANYTHING back to my siemens.com
address, please, or I'll get fired.
Thank you.
Since you started with this Star Trek analogy, I will stay within it so you
can understand what's REALLY going on with this merger.
My fellow 0-in coworkers and I did not appreciate what you wrote.
Your statement about 0-in being Borg-absorbed into Siemens MENT corporate
marketing and "left to die a slow death" shows the world that you don't
know s**t about what is going on in the formal universe despite the tons
and tons of links to formal articles.
Let me set the story straight. A lot of people think Formal ABV as a
technology is "something out in the Delta Quadrant." It is not. Cadence,
Siemens, and OneSpin are proving it.
MENT marketing may have tried to Borg-absorb 0-in, but 0-in refused to
assimilate. Most importantly, 0-in continued to innovate instead of
succumb to the Mentor corporate Borg.
You're correct in implying that during the 10 or so years after the 0-in
acquisition, MENT pissed away their lead in functional verification with
the typical arrogance of a market leader.
But formal is thriving on our 0-in island inside Siemens MENT.
By my count, the 0-in R&D team has added a shitload of new products and
significant capabilities since the original 0-in acquisition 17 years ago.
In rough chronological order (I'm pulling this off of my memory so some
items may be off, but they are all products we are selling and supporting):
2004: adding to Archer and Checkerware: 0-in checklist, 0-in Check,
0-in Confirm, 0-in Search
2007: CDC static, protocol, and metastability modeling ("CDC-FX")
2010: AutoCheck
2012: CoverCheck, Xinit Checks, Assertion Generation Utility;
Archer evolves into PropCheck
2013: Connect Check
2014: Secure Check, Register Check, X-Check, PropGen
2015: Fault injection, new Formal Verification IP
2017: Formal coverage
2018: Connectivity Explorer
2019: PropCheck deadlock analysis
2020-21: improved formal coverage, deeper integration with
verification management
I can also say (although I am not supposed to) that the annual growth rate
CAGR for 0-in formal revenue has exceeded 15% each year for every one of
the last 5 years without any acquisitions. This is beaucoup bucks for ABV.
0-in Formal had 15%+ CAGR revenue for the past 5 years
|
And you wonder what other real innovation is going on? Well let me just
start with DVCon US 2021 -- an event that will be ever-remembered as the
Jim Hogan Memorial DVCon of 2021.
Ping Yeung (Ph.D. from Univ of Edinburgh, Formal Whiz), Mark Eslinger
(leading Formal Verification Expert who makes DV guys gush when he shows
them how to become a master using formal), and Jin Hou (Ph.D. from Univ
of Montreal who is an amazing technologist with a background in finance
and formal!) -- all Siemens MENT 0-in R&D guys -- together just won the
DVCon USA 2021 Best Paper on a clever way to structure bug hunting.
This award is voted on by conference attendees. Their paper, "Formal
Verification Experiences: Spiral Refinement Methodology for Silicon
Bug Hunting" is super cool. They invented Formal Bug-Hunting Radar
Charts to guide formal users. (The PDF is here.)
By the way, I didn't see Cadence nor Synopsys win anything here; NOT 1st,
NOT 2nd, NOT 3rd place in neither papers nor posters.
Another recent Best Paper award winner was at DAC 2020 co-authored by
0-in/Siemens R&D leader Jeremy Levitt. In concert with engineers at
ARM, this 0-in veteran showed how unique new capabilities in PropCheck
enabled ARM to unearth issues in a large instruction fetch unit FSM,
get full proofs for L1 data cache arbiter logic, and find bugs in a
credit-based protocol as well as prove that it would not "lose credits".
This is mighty stuff, John!
And in just the most recent DVCons alone the 0-in/Siemens team has
partnered with customers or self-published the following:
DVCon USA 2021
-"Bounded Proof sign-off with formal coverage" -- Intel & Siemens
-"Preventing Glitch Nightmares on CDC Paths: The Three Witches"
by Mediatek & Siemens (BTW, this won 3rd Place Poster award)
-"Handling Reset Domain Crossing for designs with set reset
priority flops" -- NXP & Siemens
-"Bringing Reset Domains and Power Domains together; Confronting
issues due to UPF Instrumentation" with NXP & Siemens
- (I'm relisting this here for completeness sake) "A Spiral
Refinement Methodology for Silicon Bug Hunt" - Siemens
DVCon Europe 2020
-"Silicon Bug Hunt with Deep Sea Fishing" - Siemens
-"Improving FPGA Design Reviews Using Analysis Tools" - Siemens
-"Reliable and Efficient Resets with a Comprehensive Reset Domain
Crossing Verification Solution" - Big Fish Semi & Siemens
-"Faster Reset Verification Closure with Intelligent Reset Domain
Crossings Detection" - Siemens
-"Verify Encoder And Decoder using Formal" - Siemens
-"Using Formal to Prevent Deadlocks" - Siemens
DVCon USA 2020
-"Deadlock Verification for Dummies using SVA and Formal" - Siemens
-"Systematically Solve Reset Issues in Automotive SoCs" - NXP & Siemens
-"Are You Safe Yet? Safety Mechanism Insertion" - Siemens
-"Scalable Reset Domain Crossing Verification using Hierarchical
Data Model" - Siemens
-"Did Power Management Break My CDC Logic? An Integrated Approach to
Power Domain and CDC Verification" - Infineon/Cypress & Siemens
DVCon USA 2019
-"Modal Analysis for CDC Efficiency and Accuracy" - Intel & Siemens
You can't make this stuff up. Do you see that customers are actually
presenting how they are able to solve critical problems using 0-in's
technology all the way to today, in 2021?? Over a decade after what
you said was a "slow death." Bah!
Again, since you started with this Star Trek analogy, I will stay within
it so you can understand what's REALLY going on with this merger.
MENT marketing may have tried to Borg-absorb us 16 years ago, but the
tables were turned on them. Unable to assimilate 0-in and attain
corporate Borg perfection, MENT marketing withered away in mediocrity
in the Delta Quadrant.
In the last five years, I have seen a focus on 0-in execution that I have
not seen before. Energy levels are up -- electro-plasma is flowing well
across the starship. And there is no sign of that old MENT marketing in
any of the plasma conduits.
Finally, I take a risk here in sharing with you that Siemens had a 14
person Tiger team to assess aquiring OneSpin and all 4 of the 0-in R&D
representatives voted strongly for the acquisition. We are very much
looking forward to working with the OneSpin team to *formally* check all
8 corners of Federation space. (Yes, space has 8 corners, not 4!)
So, John, if you still want to make a Star Trek reference about the
Siemens plus OneSpin deal, I'm thinking the better reference would be to
Star Trek Voyager -- where the Maqui crew *joined* the Federation crew
on-board the Voyager -- and since you asked in your closing -- to start
giving Anirudh serious migraines in Formal ABV. ;-)
- [ Angry Siemens 0-in R&D guy ]
P.S. Thank you for letting me vent if you choose to publish this.
---- ---- ---- ---- ---- ---- ----
Related Articles:
How CDNS COVID-19 indecisiveness gave Siemens #2 slot in Formal ABV
OneSpin's "going where Cadence isn't" approach is Best of 2020 #4a
OneSpin extensible ABV plus formal linting is the Best of 2020 #4c
OneSpin DV-Verify's surprise comeback in ABV is Best of 2019 #5a
OneSpin in formal areas where Cadence is NOT is Best of 2019 #5c
Raik on it's only OneSpin vs. Jasper in formal; not SNPS nor MENT
OneSpin CEO cites 8 "insufficiencies" in Jim Hogan's Formal Guide
Hogan on Cadence, Mentor, OneSpin, Real Intent, Synopsys formal
Join
Index
Next->Item
|
|
