dac2000 panel proposals 529, 535 andrew b. kahng january 13, 2000

5
DAC2000 Panel Proposals 529, 535 Andrew B. Kahng January 13, 2000

Upload: wesley-pope

Post on 13-Jan-2016

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: DAC2000 Panel Proposals 529, 535 Andrew B. Kahng January 13, 2000

DAC2000 Panel Proposals 529, 535

Andrew B. Kahng

January 13, 2000

Page 2: DAC2000 Panel Proposals 529, 535 Andrew B. Kahng January 13, 2000

529: Dot .Com: How the Internet Will Change the EDA Business Model

• Issues– SOC/EDA issues, implications for design capability, CAD infrastructure– tradeoffs between in-house, off-site design capability– potential of Internet-based infrastructure to address design bottlenecks– technology, financial implications (selling, valuation...) of e-services model

• Perspectives and example questions– EDA: How does web help address customer needs? What happens to existing

channels? Will we see total solutions, portals, (interoperability)? Is the web a great equalizer ?

– ASIC user/vendor: How does web affect TCO of CAD infrastructure, project-based accounting? How will it affect how end users design, how vendor deploys services? Security, access issues?

– E-services solution provider: What’s real? Security/IPP/availability? Accounting? Who will buy the services, and what’s the market size?

– Investment: How will e-services affect valuations, (startup) strategies?– Naysayer/Pessimist: Hold on... here’s why it’s not going to be so easy...

Page 3: DAC2000 Panel Proposals 529, 535 Andrew B. Kahng January 13, 2000

529: Dot .Com: How the Internet Will Change the EDA Business Model

• Panel Composition

– Organizer: Tom Quan Monterey Design

– Moderator: Jennifer Smith Dain Rauscher Wessels

– Steve Domenik Sevin Rosen Funds (investment; valuation, strategy, ...)

– Bruce Toal Business-Critical Computing, HP (e-services; solns, tech, issues)

– Adriaan Ligtenberg SrVP e-/i-Cadence (Lg EDA; models/mkts, benefits/risks)

– David Dick Dir AdvTechDev, Fujitsu (ASIC use/vend; CAD TCO, usemodels)

– TBD (committed) Intel (user; naysayer, pessimist

– Jacques Benkoski(?) CEO, Monterey (Sm EDA; models/mkts, benefits/risks)

Page 4: DAC2000 Panel Proposals 529, 535 Andrew B. Kahng January 13, 2000

535: Design Closure: Dream or Reality in Deep-Submicron Designs?

• Goals– Maximize value, impact of DAC panel on this acknowledged critical issue– Distinguish competing visions without another year of marketing hype– One year since Goering’s EE Times article: What have we really learned?

• Novel structure– Panelists must agree to attempt “design closure” on a common benchmark

• NEC “G-top” design, Cisco design

– Users (NEC, Cisco) will be on the panel and discuss results as they wish– Choose panelists by industry taxonomy (not by symmetry)

• Example questions– What is scope of design closure (“timing closure”, “RTL-to-GDSII”) ?

• new metrics ?

– What are hard justifications, results, differentiators for each variant approach to achieving design closure ?

• coupling/unification? raw speed? bottom-up? right objectives?

Page 5: DAC2000 Panel Proposals 529, 535 Andrew B. Kahng January 13, 2000

535: Design Closure: Dream or Reality in Deep-Submicron Designs?

• Panel Composition

– Organizer: Michel Courtoy (Vic Kulkarni) Frequency

– Moderator: Gary Smith Dataquest

– Tera, Aristo uarch/RTL opts, commodity back-end S&P&R

– Synopsys, Cadence, Avant! more RTL-opt, std unifications, signoff P&R

– Monterey, Magma nom. structural-HDL start, unifications++, signoff P&R

– SiPerspective, Sapphire, CLK CAD (Mentor) incremental, slice of flow

– Frequency, Simplex bottom-up closure starting from signoff analysis

– Cadabra, Altius/Sagantec, Rubicad, CoreMaster/AMPS/internal

– NEC (Yoshi), Cisco (? not AVB), IBM (Reynolds), MIPS (Lev)

• Personal view: this panel has risks, but could be great