Intel Challenges AMD With i7

• Email
• Share
• Print

Intel (Nasdaq: INTC) kicks off its annual developer forum (IDF) next week. The chip giant will share lots of juicy new details on its next generation of microchips, and got the party started early with the official name for the microarchitecture formerly code-named "Nehalem."

Say hello to the Core i7 chip family. Catchy, isn't it?

Nehalem-nee-i7 will bring Intel into the age of modern system architectures, where Advanced Micro Devices (NYSE: AMD), Sun Microsystems (Nasdaq: JAVA), and International Business Machines (NYSE: IBM) have hung out for ages. The Core 2 chips, powerful as they are, work on an antiquated concept of running every processor core across a shared front side memory bus, while i7, Opteron, UltraSPARC T2, and the PowerPC 900 series have all figured out how to make better use of shared memory without penalizing the system for switching contexts too often.

If the last paragraph sounded like egghead mumbo-jumbo, just know this: The new Intel platform will handle multiprocessor systems much, much better than the Core 2 family does.

This erases another bastion where AMD was leading, with superior memory controllers. And Intel is using an advanced 45-nanometer manufacturing process (while AMD is still making 65-nm chips). It's an eternal game of cat-and-mouse, where AMD has been looking better lately, but i7 could change the game again when it's released in the fourth quarter.

Of course; that's not all there is to the i7 changeover. New power management features should please owners of huge data centers, like Google (Nasdaq: GOOG) and Microsoft (Nasdaq: MSFT). And we had to wait for a leaked i7 logo before catching wind of that development -- surely Intel has something more exciting up its sleeve than an obtuse product name.

Further Foolishness:

• As Intel's Tide Turns
• AMD's Enchanted Tiki Room
• The Best Semiconductor Stocks to Own Now