December 08, 2003
A Day in the Life
Please note that contributed articles, blog entries, and comments posted on EDACafe.com are the views and opinion of the author and do not necessarily represent the views and opinions of the management and staff of Internet Business Systems and its subsidiary web-sites.
Thursdays are never good around here. I'm always on deadline, grumpy, and short on patience. I shouldn't accept phone calls or agree to go out and see companies. I should know better.
Today's Thursday, but I went out nonetheless. I attended a small luncheon hosted by a big company for a small handful of editors. There were nice sandwiches, fresh coffee, and a selection of salads. There were nice people in a nice conference room sitting around a nice table, with the potential for nice conversation. The editors all got a nice gift when it was over. What could be nicer?
But it's Thursday. I was grumpy at lunch and grumpy all the way back to the office. I still am. The article I had hoped to run in this space isn't finished.
In lieu of that article, I'm going to share the basic flow of the conversation at lunch. The conversation around the table, before and during lunch, wandered from topic to topic - the editors discussing things between themselves, with an occasional verbal nod from our hosts. Eventually our hosts guided the conversation around to laying out some of their technology initiatives for next year. They also invited questions and feedback.
Here's roughly how the conversation went. This is the cleaned-up version. The darker version is a lot grumpier.
Why do engineers over the age of 35 often prove ineffective, or even unemployable?
They are 10 to 12 years out of school, increasingly obsolete, and usually replaced by younger, cheaper, and smarter engineers. Companies don't want to pay the price of keeping more senior engineers on board.
What are older engineers to do?
They should enroll in their closest on-line university, quickly earn an MBA, and hope they can salvage their career by transitioning over into a management track.
What happens to companies when all of their engineers are young, without a history in the industry or in the company?
tools suppliers to provide expertise for tools usage.
What was it like in the 'before' time?
It used to be that a company would encourage you to get your Ph.D., to pursue advanced technical expertise. But as companies began to go more and more public, companies began to treat their engineers more and more like Kleenex - as something disposable. The companies said their older engineers simply weren't needed anymore.
But can't companies see that they're losing a lot by letting go of their seasoned engineers?
Not really. They see that the younger engineers work 60-hour work weeks. The old guys only want to work 40. What the companies don't see is that the older engineers know what they're doing, so that in their 40 hours they're accomplishing as much, if not more, than the younger guys are in 60 - the younger guys who are still learning their craft.
Are we mostly talking about Digital Engineers here?
You betcha. Digital design is a young man's game. It's a short life - but a merry one.
What ever happened to the Product Engineer?
This was the guy who had a larger perspective on things. He (or she) had a good grasp of physics and chemistry and engineering and business issues simultaneously, and could understand the full implications of design decisions, and how those decisions impacted things up and down the design-to-manufacture-to-test-to-market chain. That guy no longer exists - fault the universities for failing to turn out these kinds of people. Today's engineers know more and more about less and less.
Product Engineers may be a thing of that past, but what about Analog Engineers?
Yeah, nothing's changed there. The analog guys still continues to do the black magic. They're still very much in demand.
What are companies in the semiconductor industry worried about right now?
Many things, particularly yield.
Because even though some people have made the move to 90 nanometers, their yields are proving to be no better, functionally, than at 180 nanometers - it costs more at 90 nanometers, but the yields are no better. Also at 10 layers of metal or more, there are so many processing steps required, there's all that much more room for error, expense, and difficulty. Also, many semiconductor companies are still waffling around between high and low-k dielectrics.
What's being done about it?
Our luncheon hosts offered - one thing that would help is to facilitate knowledge sharing across the food chain, so that Test can talk to Manufacturing can talk to Design. The editors were not starry eyed at the prospects. The editors said that there's little motivation on the part of many of the parties in the food chain - in particular, the foundries - to share process data back up stream to facilitate yield, or ease of design. The foundries often see this data as proprietary and don't perceive any benefit from sharing this information. Our hosts said that everybody wants to get along. The editors shifted around in their chairs.
Isn't packaging an issue today?
It sure as heck is - the package is now part of the design, part of the overall circuitry. There are now, and will continue to be, a host of issues surrounding package development and integrity. Packaging is hot.
So besides yield and packaging issues, what else is a big problem for customers today?
Time-to-market. Same old. Same old. All the editors around the table nodded. It's always time to market. Even though your gut instinct might drive you to conclude that things have changed in the downturn, it's still all about hitting that market window. End of story. So now you're still in a pickle. You need to hit your market window. But you also need to make your product bug free.
What's a company to do?
Kick the product out the door on time and hope that your end-user doesn't end up debugging the thing for you - or worse yet, hope that your end-user doesn't go away mad and find some other supplier next time around.
Why is this any different than it's always been - hasn't it always been an optimization problem?
Yeah, you try to find the right cross-over point that maximizes quality and minimizes time-to-market. There's nothing new under the sun there.
The topic changed - isn't strategic partnering with companies a difficult thing?
Our hosts said that partnering is now, and has always been, very important to them. Partnering is a delicate business, one that requires finesse and a mature understanding of the different nuances across companies and cultures.
How do you deal with 'partners' who refuse to show partiality towards your company's technology?
Our hosts said that dealing with partners who are aggressively neutral can be very challenging.
What are you doing in the FPGA space?
Our hosts said they don't really see any way to make money in the FPGA space. Structured ASICs, however, are a different story.
Really? A lot of people may not agree with you on that FGPA thing.
Our hosts asked the editors to explain to them how one makes money in the FPGA space.
You can find the full EDACafe event calendar here.
To read more news, click here.
-- Peggy Aycinena, EDACafe.com Contributing Editor.
Be the first to review this article