Great talk by Prof. Tedlow on the 386

Professor Richard Tedlow gave a really interesting and fun talk at the computer museum recently. You can see the lecture here or read a summery  here.   Tedlow wrote a back about Andy Grove that while very much worth reading was a bit unbalanced (I think most of his sources where people Andy gave him). My original comments at here. The book dealt with what made Andy a great man (and he is that) but not well with some of his limitation and especially those that are now reflected in the current version of Intel. This talk about the 386 and Intel’s decision to go against current business practice is a very important story in many respects. 1) It tells how Intel pretty much fell into this strategy by a series of important but tactical decisions but not with a vision of where it would ultimately lead (and I applaud Andy very much for owning up to this) 2) How the decision coupled with IBM’s previous decision to allow the DOS O/S to be kept by Microsoft and licensed freely, turned the computer industry from a vertically integrated industry where the company that had the ultimate relationship with the customer was in the power position to a horizontal structure where where two of the component manufactures (Intel and Microsoft) had most of the power and therefore most of the profits. I think there are other cases of ingrediant brands having a lot of power but do agree that this situation was very unusual. The real archetect that made this all possible was really the early pc group at IBM. Althought they did not understand the business implications.

Prof. Tedlow, shows how as a result of this, Intel went from a company that had already started to loss significant money because of Japanese Competition in the D-Ram business (the very business that Intel had created and which was its foundation) to one of the most profitable companies in the world. I was at Intel during those wonderful years of high growth and even higher profits. Frankly, I would not be sitting her in my home office in Sonoma looking at the oak trees and the grape vines if I had not been. My contributions to the Microprocessor business were indirect at best. I push the company to market to consumers and in particular did a lot of public speaking on the topic. And by taking the lead in the computer industry to create broadband networks, I probably gave the consumer PC business a longer run way but I failed to effect Intel in the most important way possible (the future beyond the Microprocessor) even thought I tried. I joined Intel in 1984 after having served as the President of Franklin Computer a company that was to Apple what Compaq was to the IBM PC (a clone), but with an important distention which was Apple’s willingness to use the courts destroy any chance we had of getting funding to support our amazing growth (almost a 100 million dollars in an our first year of operations and this was 1993/94). I use to say that when I joined, Intel was in the strip mining business and selling silicon by the ton (the memory business). Then one day the found a vein of gold (the microprocessor) and decided being in the gold mining business was a good thing. So the started extracting the gold and investing in a bigger gold mine. The only problem is eventually all the gold is tapped out. This is what has pretty much happened to Intel and to a lesser extent, Microsoft. I guess the relevant questions is what should Intel have done. I think about this still often even though it is almost ten years ago that I left Intel. Well, I don’t think there was any way to keep the top line growth and profitability long term. The demand for personal computing on a world wide basis created a kind of hyper growth or if you are a student of cosmology, it was like the inflation that may have happened in the Universe after the big bang.Eventually the growth would slow and in particular the demand for higher performance would change. Intel was in competition not with AMD but with things that other made that provided a better experience for the customer. A bigger a monitor, more memory, a bigger disc, an a nicer printer were more important than 15% more processor performance. And since Microsoft made more a more money by selling upgrades to the installed based as opposed to Intel which made its money when new systems were sold, it most important strategic partner was doing less and less to take advantage of new processor performance. It should have been no surprise that the party would come to an end. Intel needed to find a way to break out of the coffin it was finding itself.

In my opinion, the answer was there. It was on the other side of the wire in the network. But I for one could never get the company to understand this (and I tried). We did make some progress. Time Warner Cable was willing to do a deal with Intel and Oracle to form a broadband ISP (Roadrunner) and I even got an OK from Andy for us to join in this business, but are “good friends” at Microsoft found out and basically after Time Warner twice the money took over the deal (it was a stupid decision for Time Warner too since Microsoft did nothing to help them and it’s only interest was keeping Oracle and to a lesser extent us, out).  Check here.  The real problem was the Andy did not develop any company leadership that could take advantage of what Intel had accomplished find the new opportunities that continue its growth and profitability. Andy in my opinion made the worse possible decision when he moved out of the CEO role and gave it to Craig Barrett who had lead the companies manufacturing activities until he become the COO of the company. Few companies can re invent themselves. Intel did that when it went from memories to microprocesors. I guess it was too much to expect they could do it once again. And I guess companies like people have age.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s