Paul, I'd argue that the "mis-step" into the board business is too simple an view. Ask yourself what Intel's options were six months before [when the DRAM supply contracts had to be signed]. Parts were tight, but Intel could still get them. Our board customers were not all so large or well supplied. If Intel didn't agree to supply boards with memory, there was the risk that the customer would not take the board - which means they would need the processor! This was all good and well until memory prices started to fall, the customers could get parts more cheaply on the spot market and placed their board orders for 'unloaded' versions. Meanwhile, those memory suppliers, with their contracts, continued to deliver to us.
The pain we felt in Q4 and Q1 was less than the pain we would have probably felt in Q2 and Q3 had we not used our ability to get memory product. I'm sure you agree that, as an investor, you want us to ensure that those processors can continue to fly out the door!
We've said in the past that microprocessor margins have been quite stable over recent history. Our declining GM% in '95 was really due to our very rapid growth in motherboards which, no surprise, have a lower margin characteristic. Therefore, I'd say our margins contracted in '95 due MOSTLY to the influence of MIX between CPUs and OTHER PRODUCTS [boards].
Hope this helps,
John |
