forscheo,
<<I tried a single p3 vs a dual p3 500, the dual does better in heavy excel calculations. . . 1 more thing, dual p3 cracks Rc5 better then single p3>>
Obviously SMP shows some degree of performance advantage in certain applications. The question is, does it show enough to justify its cost versus a single faster chip? If you look at what you'll pay for any dual-P3 setup, you'll find a that an Athlon system of equal or better real world performance will cost less money.
Here is a very informative article on SMP written by a respected IT source that still maintains decent credibility with true techies:
<a target=new href="http://www.inqst.com/articles/p3vp4/p3vp4main.htm">
Inquest Market Research: P4 vs. Dual P3 vs. Athlon DDR</a>
The Dual P3 1 GHz is hard pressed to beat the single Athlon 1.2 GHz, and is often slower. The decisions fall about half and half to either side. 3DS Max, which has historically shown fantastic SMP benefits, puts the Dual P3-1 GHz only about 1/3 faster than an Athlon 1.2 GHz, while costing hundreds and hundreds of dollars more. And ordinary office and content creation benchmarks show the Athlon far ahead.
<<Our servers are also dual, for SQL, it's necessary, for other, it's just there, not necessary>>
Uh. . . no. It's not necessary to to have dual CPU's for anything. The only thing SMP brings you is the potential for extra performance in those applications that take advantage of it. Even if your SQL implementation provides up to 50% gains, which is almost unheard of in the industry, an AMD single-chip machine is still faster and cheaper.
Modus