The Timna wasn't suck, guys. And this is coming from a Pro-AMD guy. The Timna boasted many very interesting innovations. Since everything was integrated, the datapaths were going to be extremely large, and extremely fast. I'm almost sad they gave it the bullet. It was setting the stage for Intel to become low end, and AMD to take over the high end.
I'm really disappointed in Intel. The TIMNA was a brilliant idea in line with the Cyrix Media-GX chip. $100 PC's were within grasp. Now that they decided against it they'll move to more expensive endeavors.
Bottom line is that a 486 can do the internet. The 486 could clock at 800mHz on the .18 process, and possibly hit 1gHz on the .13 process. The chipset on the 486's was simple enough cpu, USB and PCI controllers, and memory chipset could have been pushed into a die about the size of a dime.
Intel's marketing department has pushed for all this fpu and SiMD crap to spur their sales. They've yet to enjoy their technology long enough to develop it all the way through. Rather, they push new unproven techniques at every turn. SSE and MMX doesn't enhance the internet experience. Nobody needs the fpu unit to browse HTML. Like I said before, the 486 was plenty for the internet...
I disagree with you in one regard, ColdTech. We need Intel, just like we need AMD for the competition factor. Let's face it, that's the primary reason CPU prices keep dropping. And even though I'm one of AMD's biggest fans, I'm sure their pricing structure would be very different if they were the only game in town.
I'm surprised Intel didn't just shrink the old 440BX chipsets for the Timna. With the BX's synchronous timing and Timna's penchance for being aimed at the SDRAM market, the shrunken BX chipset was the way to go. Its almost as if Intel doesn't want maximum performance out of their newest lowend processors.