(ICU 04 Item 14 ) ----------------------------------------------- [10/07/04]
Subject: What Users Would Change About Cadence
SEND LAWYERS, GUNS, & MONEY: Some favorites are stop changing product
names, stop forced bundling of products, fix licensing, fix support,
implement an open OpenAcccess, implement SystemVerilog, get an STA tool.
A voodoo priestess in New Orleans swaps your soul with Mike Fister,
the CEO of Cadence. Even though you were a Cadence customer before,
now that *you* are the new CEO of Cadence:
What 2 or 3 things about Cadence would you DEFINITELY change?
I'd never let that slimeball in my body. Gives me the willies just
thinking about it.
---- ---- ---- ---- ---- ---- ----
Licensing sucks! Hate how invoking one tool feature grabs lots of other
(sometimes unrelated) licenses. This has to change.
---- ---- ---- ---- ---- ---- ----
I would simplify the licensing schemes. I would make it easy to know
what tools you owned, what you needed, and how many it was going to take
to tool-up your design team. I'm quite positive that the complexity of
their licensing is making them millions, however. So the board would
probably lynch me first thing the next quarter.
Also I would make the documentation HTML and be done with it.
---- ---- ---- ---- ---- ---- ----
Stop changing the product names!
---- ---- ---- ---- ---- ---- ----
I would allow all tools to sold as point tools and as part of bundles.
It seems that when they have a tool that I want, I have to buy some bundle
that I have absolutely no use for 2/3 of it.
---- ---- ---- ---- ---- ---- ----
I'd change the training class food and bundling of products
---- ---- ---- ---- ---- ---- ----
Quit being a second-rate company buying tools to grab marketshare and
subsequently lose those customers because of poor service, bad marketing,
and stagnated technology. Bundling a good product with poor-selling
products and forcing people to buy the entire package at a higher cost
to get that one good-featured product is a dumb idea that just pisses off
the customer. Take the technology and push it further without killing
off the customers who bought the good tool in the first place.
---- ---- ---- ---- ---- ---- ----
Bring some of the hotline support and tool development work back on
shore. Sell the customer the tool he wants where he wants it and not
dress it up with a bunch of unneeded lipstick.
---- ---- ---- ---- ---- ---- ----
1. I would not assign critical support cases to AE's in India. AE's in
India lack the commitment to handle critical customer cases.
2. I would direct the company to clean up the multiple timing engine
problem. When Cadence acquires a tool they should commit to
implementing "their timing engine" in the tool within 6 months.
Further, the current CTE is broken. It does not time my design
correctly. FE Timing Engine times my design correctly and is
comparable to PrimeTime. I think the current CTE a weakness for
Cadence.
---- ---- ---- ---- ---- ---- ----
Customer support
Product Deliverables on time
Creat awareness in designers world
---- ---- ---- ---- ---- ---- ----
I'd fix support and their website SourceLink
---- ---- ---- ---- ---- ---- ----
The technical support. I would have the phone/email support people
properly trained.
---- ---- ---- ---- ---- ---- ----
Train all the support staff (hotline) on a regular basis. I'd also give
real job training for marketing people, have them use a tool under
pressure for a real design project that needs to work the first time.
---- ---- ---- ---- ---- ---- ----
Respond to more PCR's. 80% of all PCR's go into "backlog" status, which
basically means, "not going to do it."
---- ---- ---- ---- ---- ---- ----
- reduce PCR fix time
- open PCR database to customers
---- ---- ---- ---- ---- ---- ----
Finally getting that unified database (OA) implemented completely and
thoroughly once and for all.
---- ---- ---- ---- ---- ---- ----
Truly, and I mean TRULY, donate OpenAccess to SI2. Cadence CLAIMS to
have donated OpenAccess to the SI2 coalition, and they signed some legal
document stating so, but guess who continues to develop and support it?
And Pcells, which are integral to the OpenAccess data model, continue to
work only in Cadence proprietary SKILL. How are other companies
supposed to operate on that?
---- ---- ---- ---- ---- ---- ----
Make OpenAccess stable and a reality & make all tools working in it
---- ---- ---- ---- ---- ---- ----
get off their butt implementing system verilog
---- ---- ---- ---- ---- ---- ----
Better communication/feedback to customers about current bugs. Support
SystemVerilog 3.1a, don't be left behind by Synopsys.
---- ---- ---- ---- ---- ---- ----
Fire the VP of Simulation tools marketing guy, let the engineers loose on
System Verilog implementation.
---- ---- ---- ---- ---- ---- ----
Different levels of software with much cheaper prices on the basics or if
you need no support. Bring back ICUG conference field trips to amusement
parks or other fun destinations.
---- ---- ---- ---- ---- ---- ----
Increase R&D
Cooperate wih Sagantec
---- ---- ---- ---- ---- ---- ----
1. lower prices
2. fix backwards compatibility
3. fix releases w/o bugs
---- ---- ---- ---- ---- ---- ----
Get an STA tool that a majority of fab's will actually authorize for
sign-off. It is time to give PT some actual competition.
---- ---- ---- ---- ---- ---- ----
Get a DRC/LVS and STA tool. The acquisition strategy will consistently
make Cadence a methodology follower.
---- ---- ---- ---- ---- ---- ----
Have sales reps that understand what they are selling.
---- ---- ---- ---- ---- ---- ----
Make the sales force in analog/full custom realize the customer is not
bounty. Put more R&D money into the analog side of things, it will pay
off in the digital under 50 nm side, too!
---- ---- ---- ---- ---- ---- ----
stop buying a new compagny, sell the tool and 2 year after say "we are
going to stop selling this tool".
---- ---- ---- ---- ---- ---- ----
Understand why your top developers leave, to form startups (that you
eventually buy back), instead of working within to develop the same
tools.
---- ---- ---- ---- ---- ---- ----
Respin the services div and products to address ROI based tool use.
---- ---- ---- ---- ---- ---- ----
I'd start mimicking Magma. :-)
---- ---- ---- ---- ---- ---- ----
bring Composer out of the 70's
finally write a netlister that works
reduce simulation elaboration memory footprints
---- ---- ---- ---- ---- ---- ----
Improve quality of the releases, more flexible eval license.
---- ---- ---- ---- ---- ---- ----
When CDN makes a claim for the capabilities of their tools, it's never
clear if this is a currently released capability, or one that is expected
in the next release, or one that is being discussed and might be put
into the tool. I would make tool roadmaps/features more visible.
---- ---- ---- ---- ---- ---- ----
Fish or cut bait with OrCAD. Quit sucking the profits out only to put
out an inferior update like OrCAD 10.0 -- now on it's 2nd service pack in
as many quarters (and it's STILL buggy -- I'm waiting for SP3). I'd
either turn it loose and move on or bring it back into the Cadence fold.
---- ---- ---- ---- ---- ---- ----
I would no longer destroy tools from acquired companies (e.g. Signalscan).
Improve the "ease of use" of tools. Set-up is confusing and frustrating.
Enhancements to make debug easier (e.g. searching for signals, modules,
nets). Interface with 3rd party tools if in-house tool is inferior.
---- ---- ---- ---- ---- ---- ----
put more effort into the PCB tools; make the router a lot more functional
|
|
