advertisement
Click Here. 
To print: Click here or Select File and then Print from your browser's menu

	--------------------------------------------------------------
	This story was printed from Anchordesk,
	located at http://review.zdnet.com/AnchorDesk/.
	--------------------------------------------------------------

Athlon 1.2 GHz Visited -- P4 Revisited
By Bill Machrone: Contributing Editor PC Magazine
Wednesday, November 29, 2000
 

Few topics spark more controversy than the introduction of a new Intel processor, especially in the context of the ongoing competition with AMD.

A lot of people weren't happy with our coverage of the P4 because we didn't include results for AMD's new 1.2 GHz Athlon. The Athlon was missing from this particualr review because we didn't have a comparable machine in house -- our last 1.2 GHz machine had RAID drives.

The howls of outrage from online readers were something to behold. They variously claimed:

In fact, the only thing that's missing from the Talkbacks is a "Mac rules" message -- they usually come out of the woodwork whenever the talk turns to X86 processors.

Well, if you're curious, we've got more info. Go back to the Performance Tests link and you'll find comparisons to a white-box Athlon 1.2 GHz clone. And yes, the Athlon did better than the P4s on most of the tests.

But first, a comment about our esteemed readers' comments.

First off, benchmarks can prove whatever you want them to prove. Some other sites show the P4 superior to the Athlon 1.2; others show the Athlon kicking Intel butt. In both cases, however, the difference in performance is undetectable to the human eye. The difference between the two won't get you home any earlier.

Second, most of the messages represented what we used to call "bench racing" in my automotive days -- the theoretical advantage conferred by various engine modifications, suspension setups and so forth. But we all knew that the driver's ability made far more difference than some subtle tweak under the hood.

Computers aren't so very different -- what good is all that processor speed if you can't type, spend hours looking through the help files to figure out how to use some Excel function or are crippled by a noisy, dial-up line? Most of the complainers won't buy a P4 or an Athlon 1.2, but they'll make lots of noise about the one they would buy.

In fact, both the Athlon Thunderbird (the core used in the 1.2 GHz chip) and Pentium 4 represent significant advances in the state of the art. Is Intel's solution devoted to winning the specsmanship race -- more gigahertz but less throughput? We'll never know, but the picture that emerges from the benchmark tests is that AMD's solution cranks out more results per clock cycle than Intel's. The efficiency-lover in me loves that.

Is Intel pandering to the uninformed consumer's desire to buy raw clock speed rather than throughput? If so, it's a dramatic reversal from the days when the company was trying to kill clock speed as a measurement, to shift the metric to various throughput benchmarks instead. Their reason at the time was the introduction of the 486, the first superscalar microprocessor, which performed more than one instruction per clock cycle.

At the time, a 486 running 25% slower than a 386 or 386-class processor could do just as much work, and Intel didn't want them to be judged on megahertz alone. But now that superscalar processing is commonplace, maybe it's just a matter of marketing.

We at PC Magazine apologize to our ever-demanding readers for not having the Athlon results available last week, but I'm sure it was worth the wait for you Athlon fans. Heck, even if you can't see the difference, it's some kind of moral victory.

Read PC Magazine's review of Pentium 4 Machines -- plus Athlon 1.2Click for more.