Intel Chips For The Near- And Semi-Near Future
Brian writes "This
article reports that Intel will release new chips at the Comdex
trade show, its first low-power designs for super-thin servers. The
new Pentium III model is a
gussied-up chip taken from the company's product line for portable computers,
which share many of the same constraints as ultradense
servers. These systems can't consume as much power or give off as much heat as
ordinary CPUs because overheating causes processing errors. The systems
are the first swing of a one-two punch against Transmeta,
whose low-power designs caught Intel
flat-footed, first in the mobile market and then in the low-power server market. Intel now is fighting back just when most
server companies using Transmeta chips
are on the ropes." And albat0r writes: "Intel says that it will hit 3GHz on the mainstream Pentium 4 by the end of 2002. Intel will advance its Celeron line, currently based on Pentium III technology, with Pentium 4 technology by mid-2002." I look forward to good values on eBay when 2GHz is "obsolete."
Low power is great, there is a definite need for using less power and producing less heat in some systems. As for MHZ increases, I truly wonder what is driving the need for speed anymore other than media types and gamers. Where are the next generation apps that will utilize this kind of firepower? Media producers, avid gamers, engineers, and server roles excluded, who else needs or even wants this kind of power? What will you do with it, besides *everything* you do today "faster"?
Wanna get high?
I think the P4 was an especially telltale sign of the times ahead for the PC chip industry. While the AMD rivalry has helped to spark fierce pricing competition, I also think that it has prompted Intel to go the "MHz at any cost" route. Don't be suprised to see the P4 as the first in a long line of "let's increase that pipeline to pump up the clock speed!" While undoubtedly this can make a chip faster if IPC is not just cut equivalently, it also smacks of "MHz at all" marketing strategy.
Intel has realized (more than AMD, who is still trying to "educate" those consumers) that the general mass of people don't pay attention to SIMD instructions, double clocked FPU units, superscalar speculative execution, full speed caches, or any of that other jive that gives you higher IPC. They look at MHz and just want to see higher numbers. And also more CPUs can't = bad either, can it? I mean, that's the next marketing blitz campaign once MHz stops working.
Agreed. Intel seems to be missing the boat with Itanic. Constant delays, (we've been hearing about Merced since the mid nineties,) a complete lack of interoperability (I personally think that's a GOOD thing, but it makes it harder to sell,) and a ridiculous pricepoint (will eventually change, of course,) intel is missing out while AMD forges ahead with its consumer level product.
Keep in mind, though, that you will not be able to use 64-bit addressing in 32-bit compatability mode with the AMD offering... you'll still be stuck with the 4 gig limit when running "legacy" code.
If both products were available today at the same price, and if applications were available, I'd have to go with the Itanic. Intel has thrown away the last 20+ years of legacy crap that has been slowing development for so long. AMD, in an effort to ease migration and appease the masses of consumers, has retained those old roadblocks and built new roads around them.
AMD also doesn't have the marketing to push their product and make it mainstream. Remember, in the modern economy it doesn't matter if your product is better: people have to BELIEVE that it's better. AMD isn't good at this, intel is.
Low power CPUs certainly can help drive down the costs in less than obvious ways. With a CPU running at maybe 20W less power - a data centre of 500 dual processor servers will be saving an incredible amount of power - both in terms of electricity required to power and aircon to cool. My bad math says maybe 20kW, 24x7...
Evil ZEN Scientist