1990-03-10|Sunatori's Observation (Towards a Better Corporation: The Legendary Years)

Philosophy, Principle, Value and Mission Statement

• contribution to Canadian society through science and technology
• kilometre-post in BNR/NT CMOS4S/CMOS4 development (focus)
• worked for BNR, not for any person
  • professionalism and respect for human dignity
  • inspired, motivated, led and supported by Stephen Sunter

Long-Term Vision, Direction and Foresight

• always think forward the next step, the next time
• bear short-term pain for long-term gain
• do permanent solution at the source instead of endless quick fixes
• do quick fixes with the understanding and commitment to do fundamental root-cause fix later
• do root-cause analysis
• embrace and be prepared for changes
• never compromise on quality over quantity and schedule
• keep master plan up-to-date
• use lateral thinking instead of linear thinking
• corporate framework with long-term vision (VHDL standard, Configuration Management)
• envisioned digital-only processing (A/D - processing - D/A)
• envisioned fully synchronous Finite-State Machine-only design system
• envisioned Multiple-Level Metal interconnect and stacked vias
• Cell Library technology demonstration plots (follow Technology Road Map):
  • 1985: CMOS Technology Evolution
  • 1985: CMOS4 Standard Cell Library
  • 1986: CMOS4/CMOS3 ZIPPER Cell Library
  • 1986: CMOS4 800K-transistor (133K-gate) Channelless Gate Array
  • 1987: CMOS4S Technology Capability (Triple-Level Metal, AOI/OAI)
  • 1988: CMOS4S Technology Benchmark (3-bit Cruise Control FSM)
  • 1989: CMOS4S HyperCell Library

• electronics ---> optics (optical Integrated Circuits with GaAs, InP)
• exploratory research and development
• always check historical background

Absolute Simplicity and Ultimate Consistency

• "good design is simple design"
• error-cause removal (single source versus redundancy, copying, comments)
• make system dynamic, generic and universal (completeness)
• don't hard-code, but always soft-code and soft-link (technology independence)
• correct-by-architecture method for simulation-less design
• quality through correct-by-construction method, not by checking and testing
  • not statistical but exhaustive testing only

• Total Quality Management System with absolute simplicity and consistency
• 0 error tolerance policy instead of Statistical Process Control
  • absolute with controlled exceptions (ATM/SONET deficiency)

• correct-by-construction instead of checking and inspection
• minimum number of cells, files, programs/routines, lines of code, documents
• the best documentation is no documentation at all
• perform traceable fundamental, lasting, system-wide changes globally to lower total maintenance cost
• not local but global optimisation for scalability and expandability
• turn problems into opportunities by simplification and consolidation
• record all exceptions
• no order-dependency
• avoid overlap, duplication, redundancy
• reuse
• MISC versus RISC versus CISC

Design Process and Methodology

• emphasis on the process, procedure and method, not the product
• try to perfect data rather than tool for corporate knowledge accumulation
• close the loop first, then optimise (repeatable end-to-end test)
• idiot-proof push-button hyper-automation (CSIM, CFUNS, CDESC, CPIN, ZIPPER)
• single-entry on-line documentation access over paper manual (CELLHELP)
• modular evergreen products to manage changes by evolutionary progress
• control design margin
  • allow 0 margin and then add margins
  • don't margin margins

• document and constantly review past/present/future assumptions
• create and keep universal checklist and follow exactly
• ran DRC on all-cell demonstration blocks
• pushed for overall system cost reduction with scan-path DFT
• single parameterised model with common hardware, software, transport
• creativity in assembling knowledge effectively, not with different words (publish or perish)
• don't hide the "anomalies"; seek out and eliminate their causes
• Total Quality and 0 Defect concepts in Policies and Procedures
• overcoming technical threshold (2 approaches)
  • face the hurdle every time
  • solve it once slowly, then use the process, procedure, method for other problems

• adopt CDS paradigm: data-centred, not application-centred

Productivity and Effectiveness

• put emphasis on building solid design infrastructure first
  • implementation is easy when the design is sound and well understood

• obey strict release mechanism (CONFIG, LIBPUSH, LIBPULL, VERIFIED)
• develop general-purpose utility programs (SC, FTP_LIST, MYPLOT, PRT)
• maximum use of home terminal and parallel operations (HyperFlex hours)
• synergy
• Open Corporation
  • virtual projects
  • dynamic corporate vision technology road map
  • dynamic financial barometer
  • dynamic product promises

Standardisation and Decentralisation with Strong Central Coordination

• use only SI (metric system)
• use ISO date and time format (BNR corporate policy)
• standardise and uniformalise first, then customise (controllability) (e.g. alphabet)
• do major incompatible changes only when there is quantum-leap improvement (Record/Compact Disc versus VHS/Beta)
• use strict version control (use of PLS DEX, CHIPCOMP)
• flexible programmability and configurability with Gate Array (HyperCells, ZIPPER cells)
• set up project disk (use of LIBCNTRL, LBC, NU on CELLDB 191)
• stick to single benchmark circuit (3-bit Cruise Control FSM)
• YPM/SC4/GA4 Process Monitors
• use of 5K programs to accumulate and spread knowledge
• corporate policy on computing platform (both hardware and software)
• open architecture for transparency of integration and functionality

Ownership and Accountability

• individual rights and freedoms
• respect for engineering dignity and individual rights and freedoms
• never rely on "human touch" (late night phone call?)
• don't trust anybody but yourself
• career cell count: 490 CMOS4S/CMOS4 cells, 96 CMOS3 cells
• test-chip contributions (Y13, Y23, Y29, Y33, Y42, Y44, Y56)
• automated Cell Library builds (BUILD EXECs)
• technology file support (TECHDAT, TECHFILE, wide centrelines)

Technical Justice

• make logical decision based on technical merit rather than political one
• make policies to benefit all instead of favouring specific customers
• don't violate the rules, but fix them if they don't work
• don't be afraid to speak out against unfair and unjust practices
• don't use DESBUF, FIFO, use FINIS (close files), CLEAR (clear memory)
• avoid spaghetti code (don't use "GoTo" statements)
• avoid hard coding
• make parameters mutually exclusive
• half-grid cell battle (1985)
• millimetre versus mil conflict (1985)
• drawn gate-length versus effective gate-length controversy (1985)
• number of transistors versus number of gates struggle (1986)
• removal of untestable cell schematic on documentation layers (1986)
• gate utilisation-ratio fight (1989)

Corporate Knowledge/Bible

• Cell Library group as key "Centre of Excellence"
• establish integrated up-to-date corporate knowledge-base (DQINFORM?)
• prefer shared database to countless reports and COCOS messages
• don't re-invent the wheel again and again
• execute direct corporate communications/announcements (snowstorm memo)
• install corporate electronic suggestion box (RFA?)
• do effective information processing and knowledge build-up
• more importance in data and macros as corporate asset than CAD/CAE tools
• version control and configuration management system

Innovation

• Break paradigms
• HyperCells (CMOS4S)
• ZIPPER cells (CMOS4S/CMOS4/CMOS3)
• Modular Standard Cells (CMOS4S/CMOS4)
• ZIPPER-VELCRO, Super-ZIPPER, Enhanced Super-ZIPPER, Hyper-ZIPPER, Wonder-ZIPPER, Refined Wonder-ZIPPER, Magic-ZIPPER, Improved Magic-ZIPPER developments
• mixed ZIPPER and Standard Cells in AUTOLAY
• explored theatrics in presentations
• explored legalese in technical reports
• explored use of holograms for marketing (Good Ideas program, Innovation Centre)

A Recipe/Formula for Success

• follow the 10 points (ignore Go Simon Sunatori-specifics marked in italic)
• never follow old adage "If it ain't broke, don't fix it!"
• follow new golden rule "Fix it before it breaks!", "If it ain't broke, improve it!"
• never solve a problem that can be eliminated!
• follow some of Dr. Edwards Deming's 14 Points
• follow some of Sam Walton's 10 Rules For Success
• strive for perfection (passion for excellence, enthusiasm, commitment, competence, esprit de corps)
• try to do it right the first time (remember Murphy's law)
• do what you say you will do
• make constant and continuous quality improvements (BUGS, PRS, FP, RFA)
• follow the gate-review process and look at the total picture
• fix the cause of the problem, not the symptom (continuous cause-and-effect analysis)
• don't damage control, but predict crises/disasters/catastrophes/fiascos
• recognise honest exposure of truth to learn from experience (both failures and successes)
• foster decision-making by engineers
• say "no" to elegance, image, appearance, "yes" to substance
• alway go straight to the source with direct access, eliminate intermediate processes
• 0 tolerance for mediocrity
• "Life Management Skills for Success: 6 principles of personal management" (Lory Fischler)
  • Productive People Don't Manage Time, They Manage Themselves
  • Simplify, Don't Multiply
  • Control Tasks So They Don't Control You
  • Define Your Objectives
  • Allocate Time Based on Priorities
  • Find Something You Love to Do

• check the Gartner Hype Cycle (visibility versus maturity) (technology trigger, peak of inflated hyperbole, thought of irrelevance, slope of enlightenment, plateau of permanent annoyance)
• check the S-Curve (innovators, early adopters, early majority, late majority, laggards)

Vicious Attacks on Management

• stop managing, start leading (self-management)
• observation of the "Peter Principle"
• don't filter/screen/distort information
• don't change for the sake of change without real progress
• don't make decisions based on today's snap-shot, consider long-term impact
• listen to constructive criticisms and suggestions in context
• experience and analyse end-user frustrations
• lack of hardware/software integration (duplication and proliferation of data)
• lack of stable data format (CALMA, 16-bit/32-bit ICDATA, CDS, GDSII, EDIF)
• "magic tools will solve the whole problem" (DEX, CDS, workstation)
• vapourware (FUNDES, CBITS)
• too frequent management changes (P861, P843, 5K43, 5K54, 5L26, 5S54)
• poor communication between BNR and NTE
• initiated numerous COCOS wars and conflicts for positive discussion
• demanded usable EXTRACT and CELLIO-entering programs
• DEX nightmare led to "Cell Library Build Tragedy" report (1986)
• SDS-mover difficulties led to "Cell Library Responsibility" report (1987)
• CDS troubles led to "CDS DEX Conversion Epilogue" report (1988)
• originated gate-isolation patent infringement scandal (1989)
• unable to convince the organisation on many issues (discontent)
• inconsistency in production weeks between BNR and NT calendars (1988)
• pressured Cesar Cesaratto for better communication in Division 5 Lab Council
• demanded opening up the secret decision-making process for more democracy
• articulate alignment of core purpose, please (integrity, people empowerment and trust)
• internal communication model
  • CMOS4S performance increase example: MSS 100% -> D-level 50% -> C-level 25% -> B-level 10% -> A.V.P. 5% -> V.P. 2.5%
  • each level of management is 90% perfect * 7 levels = 48% reliable overall

• intolerance for managers who claim the Core Value of "We Are One Team" where the Team is his/her department
• intolerance for managers who show no respect for "Simple Design is Good Design"
• "peer review" means managers judge subordinates?
• "dual career path" means certain managers can have both management and technical titles?

Non-Monetary Rewards

• never claimed overtime pay
• achieved 2 patents
• received 2 BNR Awards of Excellence (1986, 1989)
• dinner conversations with BNR Presidents John Roth (1985), Gedas Sakus (1989)
• featured in Telesis (1988 one) and Insight (1986, No.5) articles
• nicknamed "King of CMOS4", "Mr. HyperCell", "Creator of SC2FM"
• too much singing, arrogance and sarcasm
• excessive self-confidence and finger-pointing
• very tough but fair on customers and colleagues
• didn't socialise with co-workers
• favoured technical superiority over pleasing people
• chose to be a champion instead of a manager
• pursued hologram sticker case through Good Ideas Program and Innovation Centre
• had great fun and satisfaction with CM4NEWS
• proposed "Silicon Oasis" for the contest to name the cafeteria
• created BNR light bulb joke
• created BNR World-Class Design System and Management joke
• received a personal memorandum from NT Chair and CEO, Paul G. Stern (1993)
  • "it is people like you that make Northern Telecom a great company"

Hypocritical Slogans

• Our Policy: We will deliver defect-free designs
• Customer is number ONE, Quality is number TWO!
• Simplification
• Effectiveness
• Innovation
• Globalisation
• Excellence!

5 Innovations

• BNR first to produce a farewell book
• Yellow (non-white) colour paper
• Intentionally round book first in the history of civilisation
• Round cover photo
• Igloo on the back yard

HLPCOM/CELLHELP/CM4NEWS

• full-screen on-line search system
• integrated knowledge base system
• hypertext
• CodedoC (code + documentation)
• single point-of-entry
• bulletin-board system
• electronic voting
• electronic democracy
• groupware
• workflow automation

Vintage Russ Berrie & Co. Quik Quips Tiles

• Do you want to speak to the boss or to someone who knows what's going on?

• ESTABLISHING PRIORITIES
  • URGENT: Get someone else to do it right away.
  • RUSH: Try to look at it before Friday.
  • ASAP: Whenever.

• 6 PHASES OF A PROJECT
  1. Enthusiasm
  2. Disillusionment
  3. Panic
  4. Search for the Guilty
  5. Punishment of the Innocent
  6. Praise & Honours for the Nonparticipants

Go to home page/Aller à la page d'accueil

Go to Weblog/Aller au Weblog

English / Français

Comments/Commentaire

"Sunatori, Go Simon" <GS.Sunatori@HyperInfo.CA>

[Donation/Don, SVP]

Copyright/Droit d'auteur © 1995-2024 by/par HyperInfo Canada Inc.

• All rights reserved./Tous droits réservés.
• Produced in Canada./Produit au Canada.

ISBN 978-0-929105-04-8

• 2024-12-17T20:35:58.000Z