The outsourcing problem is a real mess, and it is a complicated situation. Top management has large incentives for quarterly performance. One sure-fire way to improve that rapidly is to cut labor costs. Even if the design quality is below what is needed, and staff back in the U.S. have to spend extra time reviewing, correcting mistakes and retesting, the short-term effect on stock price is positive. In addition, the board of a corporation has a legal responsibility (in the U.S.) to their stockholders to maximize earnings by any legal means. Management is behaving exactly the way they should, given the structure of their incentives and their legal responsibilities to shareholders.
One could argue that in the long run, the company would be better served to develop a staff that knows the company's goals, knows it internal processes and can get a design right, or darn close to it, the first time. That requires a long-term investment today, and probably five years before you'll see much return. With stockholders emphasizing short-term performance, and the fact that most CEO's can't expect to be in their positions for that long, it would be absurd to expect them to act differently than they do. Interestingly, Intel's technology center in Bangalore is an example that combines both short-term and long-term views, and shows why that doesn't happen here. ASIC design is a particularly expensive endeavor. What Intel is doing in Bangalore is to develop exactly the type of design infrastructure I described but with the specialized labor only costing 10-20% of what it costs here. That huge gulf in wages, along with the generous support of the Indian government, allows them to make a long-term investment with relatively low risk. Simply put, their board would not likely approve such an investment in California, Oregon, Arizona or Texas due to the much higher cost and the resulting longer time to recoup their investment. However, the dirty little secret is that this is chump change when viewed from their overall cost of operations. As this is mostly a software list, I'll risk telling you something you may already know: it takes USD$3-5 billion (5e-9) and around four years to bring a new silicon foundry online. A mask set (tooling expenses) for a submicron IC starts at USD$500K and often tops USD$1M. Processors sell for tens or hundreds of US dollars each. The design labor is not a large percentage of the total product cost. So why all fuss to squeeze every last bit of cost out of it? I believe there are two primary factors. One makes sense in our capitalist economic system, but the second does not. The obvious and defensible reason is the imperative to reduce costs that I mentioned above. That is the result of our economic system and the operation of the stock market. The second reason has nothing to do with economics and is therefore harder to remedy. Top management of our largest companies, usually lawyers or accountants today, personally believe that technical personnel DO NOT DESERVE the same compensation as managers because they are not as important to the success of the enterprise. The results of that view works against their own economic interest, but like most prejudice, it is neither borne of knowledge nor wisdom. One mid-sized electronics company I worked for got a new CEO who was an accountant from the construction industry. He had no concept of what the technical staff did nor what we contributed. It really bothered him that engineering managers and many engineers had private offices with windows that looked out over the beautifully landscaped corporate campus. We felt the working environment was a significant bonus that kept many of us loyal to the company. The CEO felt this was a wrong that he had to correct. He then spent USD$1.5M to buy cubicle furniture and moved the engineering staff to interior space in the plant. Because the company had no further use for the offices with windows, they were filled with cardboard boxes and used for storage. I don't have to tell you how this affected morale. This ridiculous waste of money is an example of the deeply held prejudice of many people who make financial decisions today. They know they need technical talent, but consider technologists as a commodity. In contrast, they view people with MBA's as smart, hardworking and deserving far more than what the company can pay them. My personal belief is that this worldview is more responsible for our current outsourcing woes than the original economic reason. Here's a further example of this bias. Once as an engineering manager, I reported to a marketing product manager. When I asked my new boss why the marketing staff, without graduate training and little experience, was paid considerably more than my engineers, some of whom had PhD's and decades of experience, here's what he told me. "Management personnel are rewarded in expectation of achievement, while technical personnel are rewarded only after demonstrated achievement." While my blood boiled, I paused to think, then asked why. He replied, "I guess it's harder to find good managers than good technical people." He honestly and sincerely said those words to my face. Believe me, I couldn't make up something that stupid if I tried. Please don't make the mistake of ignoring what we are dealing with. We can't fight this prejudice with our eyes closed. -- Seth Goodman > -----Original Message----- > From: Weissgerber, Tom L [mailto:[EMAIL PROTECTED] > Sent: Friday, November 11, 2005 10:43 AM > To: debian-user@lists.debian.org; Weissgerber, Tom L > Subject: Request to remove Information > > > Debian, > The following information should not have been made available > to the entire public domain. Please remove the following > links/files at your earliest convenience. > Date: Wed, 24 Sep 2003 10:57:42 -0700 > Message-id: > <[EMAIL PROTECTED]> > In-reply-to: > <[EMAIL PROTECTED]> > Message-id: <[EMAIL PROTECTED]> > Old-return-path: <[EMAIL PROTECTED]> > References: > <[EMAIL PROTECTED]> > Reply-to: [EMAIL PROTECTED] > > > Regards > > Tom Weissgerber > Intel Corporation > Validation Tool Development Manager > 916-356-5339 > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]