although the rate of growth will likely slow down this year, according to IDC...
This is where the analysts miss the boat. I am pretty bullish on AMD for this reason. AMD has proven they have a great Opteron product. 4Q results show that in spades, no discussion needed. Intel has huge revenues in Mobile and with Yonah, I think they will keep most of them, but the lack of 64 bit and its premium pricing do make it vunerable. On desktop, AMD will eat a lot of share, they have the capacity, consumers and businesses love them, the lawsuits prevents Intel from doing anything crazy.
Now we get to 3Q and the Next Generation Architecture. If it is on time and as fast as claimed, it will greatly slow AMD's advance in units and asp's. Some analysts miss that (the pro-AMD ones), some like ML's Osha get that, but what they all seem to miss is that if this product is so great and AMD product is also great (see 4Q earnings for proof), that those great products will INCREASE the market size! Analysts like to think of the size of the pie as fixed and it is just a matter of who gets more share. The truth of the matter is with great products, x86 is growing for 3 major reasons.
#1 They are taking huge amounts of share from RISC in the server market. This alone could double the x86 pie over the next 5 years (for both Intel and AMD to share).
#2 The emerging markets are growing and the billions of people that live there will soon be able to afford at least a $100 computer and many of them a $500 computer (which is a high end desktop now!)
#3 This is the one everybody misses. If computers are a lot better than before. Witness the traditional 15% to 20% improvement per year in mainframe price performance and compare it to the 40 to 50% inprovement going on in the x86 market. Does anybody else remember the 4 way Intel machines in 1999 (600 mhz Xeons) and how long it was before Intel made a part that was better (the 1.5 mhz P4 based Xeons were barely faster for server work, so it was about 3 or 4 years that those folks had no upgrade path). As businesses move over to x86 machines, it changes the economics of the IT shop. There are all kinds of labor vs. hardware tradeoffs that IT folks like me make every day. I'm working on a medical database now and I just spent a week tuning the database because it ran too slow. We are on an old Sparc machine. If we moved to a faster x86 machine, I would have been able to add new functions (very much needed), instead of just tuning the database. I'm saying that IT expenditures go up like 10% a year, but if the hardware is better, they can substitute a bit of hardware for people. What most IT managers will do (not wanting to lay people off, because that is hard on their families), is just spend the same on hardware as last year (executives are thrilled, because they expected costs to go up) and get 10 times the capacity they got in the old RISC days. So in the short term this is really just point 1 and 2, x86 will sell more into the server and worldwide markets. But in the longer term, the new businesses (like Amazon.com, Ebay, Yahoo, Google, Biotech firms (I don't know the names), Oil exploration firms, etc.) Computing power being more affordable makes certain business models possible that just weren't possible before. Nobody could have built an Ebay in the 70's. Even though geeks had BBS's back then, you didn't have 3 things. Regular folks didn't have computers. Those that had computers didn't have access to a central network. The networking that the computers did do, was based on primative standards like xterm, so you couldn't do the easy to view web pages without installing special software and that is way too much work for normal people. Anyway, the point is improved performance per dollar will drive new uses that nobody can even imagine today.
So my overall point is that if Intel's NGA succeeds, the market will grow so fast, there will be enough business to keep both AMD and Intel happy for a few years (think Apple's IPod vs. MP3 players, they both appear to be "winning"). If Intel's NGA is not successful, Intel will be forced to drastically cut expenditures or it will spiral into financial losses. AMD is growing quickly and it may take them a few years, but if Intel doesn't come to the table with a competitive design, AMD will take enough revenue from them that they won't survive with their current cost structure.
News 
Market Data 
Discover 
AI
