been wondering this for so long -- difference between these processors

[bk12321]

I have seen Pentium M, Celeron, pentium 2, 3 and 4, athlox xp, sempron, amd 64, and a lot of other processors. I was just wondering why a pentium M 1.6ghz is more desirable than a Celeron 2.2ghz or why, for example, an AMD 64 runs at 2.0 ghz but it comparable to an intel processor running at 3.2ghz? I am no computer noob, but this is just one of those things I never picked up by reading articles or asking people, so im asking now . How come a pentium M 1.8 laptop is supposively good? I mean Athlon XP's hit clock speeds that high, but they are not as good. Also, for example you see Athlon XP's hit 2 ghz, but 64's are around the same and are so much better. Why? This has bothered me for a while and I never really tried to find an answer.

 

[pigpen]

I have seen Pentium M, Celeron, pentium 2, 3 and 4, athlox xp, sempron, amd 64, and a lot of other processors. I was just wondering why a pentium M 1.6ghz is more desirable than a Celeron 2.2ghz or why, for example, an AMD 64 runs at 2.0 ghz but it comparable to an intel processor running at 3.2ghz? I am no computer noob, but this is just one of those things I never picked up by reading articles or asking people, so im asking now . How come a pentium M 1.8 laptop is supposively good? I mean Athlon XP's hit clock speeds that high, but they are not as good. Also, for example you see Athlon XP's hit 2 ghz, but 64's are around the same and are so much better. Why? This has bothered me for a while and I never really tried to find an answer.

check the stickies in the Intel and AMD forums. Short answer is architectural differences between different processors. The workload of a CPU has two components. Clock speed and the information it can process per clock cycle. A Pentium M / AMD is comparable to a Pentium 4 with a much higher clock speed because they're capable of processing more information per cycle. And maybe even more to the point, they are much faster to recover from mistakes

 

[bk12321]

so a pentium M 1.7ghz laptop is actually pretty good, even though the actual clock speed is almost half of a pentium 4?

 

[Scali]

Think of a CPU as a conveyor belt-style assembly line.
The clockspeed is the speed of the belt.
The faster the belt runs, the less time people have to construct the product. So you need more stations along the line, with each person doing less operations.
Now if one person would run out of parts, the entire belt will have to wait on that person to continue again. The longer the belt, the more stations have to wait, and the more time is lost.

Whether faster belts or less stations are a better solution depends on the situation you're in. If you run out of parts all the time, it's smarter to just keep the pipeline short. If there is a good supply of parts, and the belt never has to wait, the longer the better, because the speed of the belt is directly related to the amount of finished products you can put out in a given timespan.

Processors work in pretty much the same way, except the conveyor belt is called a pipeline, and you're not building a product in stages, but processing instructions in stages, and each station is called a pipeline stage.
Whether the situation of 'running out of parts' (say a cache-miss) occurs often or not, depends on the application... That's the catch... there are different CPUs and different applications, one CPU is better at one application, another CPU is better at another application. There is no CPU that is better at everything. And by redesigning an application, you can make it perform better on certain types of CPUs. So a CPU being slow in a certain application could also have to do with a design that is inefficient for that CPU, not just that the CPU in itself would be inefficient at performing that task.