Both AMD and Intel claimed they gained market share in Q4 2006 at their recent analyst conferences (See AMD; Intel). According to Mercury Research, the results favored AMD, but there were some boasting rights for Intel as well.
Overall AMD increased its market share, breaking the 25% barrier for the first time ever, with 25.3% of shipments. AMD's gains came in processors for desktop systems, where it had 29.1% market share, and in notebook processors, where it moved up to 19.4%.
But in the higher margin server market AMD lost market share sequentially. In Q3 it had peaked at 23.6%; in Q4 it came in at 22.2%. Still, that was up from Q4 2005 when it had only 16.4% of the market.
I am actually surprised by how well AMD held market share against Intel's new chips that finally caught up to AMD's technology lead. Intel is still the default decision for most IT professionals; some of Intel's problems earlier in the year were because buyers were waiting for the new Intel chips rather than switching to AMD. Intel's media hype was extreme, and now that Intel chips are in Macs the Mac-biased media types weren't in attack mode.
For short term investors neither Intel nor AMD is a good bet. Both could make significantly higher profits if they would back off this market share war, but neither will. AMD is determined to keep gaining market share; Intel is determined to put AMD back in the under-10%-of-market box that allowed Intel to become fat, lazy, and stupid yet very profitable in the late 1990's.
AMD has picked all the low-hanging fruit. The major computer makers IBM, HP, Dell, Lenovo and Sun all use AMD processors now. Those IT professionals who did not buy AMD when it had a clear performance lead now have an excuse to go back to the safe choice. Intel still has a lot more money than AMD and therefore can win a price war.
For AMD the question is: how good are the new Opterons coming out in June or July going to be? Can they make Intel look stupid again? There are some hints that they can, but Intel engineers are working overtime as well, as anyone would trying to keep their jobs after the massive cuts Intel had to make in 2006.
For long term investors the best strategy is to hedge the bet by having both Intel and AMD stock. Some day they will back off their price wars and profits will skyrocket.
But the smart money is on AMD, not because it is a clear winner, but because it had so much less to lose and so much more to gain. The stock is selling for not much more than when it had far lower market share. It is a risky stock, but if AMD does reach 50% of the market in 3 to 5 years, it will be set up to be a profit powerhouse.