Yes, it's very clear that an athlon does more per cycle than a P4, which is due to the fact that the P4 has a longer pipeline from what I understand.  But this becomes moot when intel are releasing processors at over 2.5 GHz while the fastest athlon is at 1.67 GHz (the 2000+ model).  A 150% advantage in frequency more than makes up for the efficiency of the athlon.

Also, it is no longer true that a P4 system costs much more than a comparable athlon.  A quick search on pricewatch.com shows the athlon xp 2000+ (1.67Ghz) going for about $260.  A 2 GHz northwood can be purchased for under $200, and even the 2.2 is only about $240.  And now that there are viable DDR platforms for the P4, memory doesn't cost any more than it does for the athlon.

all this talk about the pros or cons of overclocking, to do what?  If you have a high end cpu what are you running that needs the extra mHz?  To have faster quake3 or photoshop time demos?

Your smoking too much pot or something, you can get a Boxed AMD 2100+ for $196 on googlegear, and it comes with a free T-shirt, so you can advertise for them for free as well.

They have a 2.2 P4 (which probably still won't benchmark as high in games) for $240.
The next step is a 2.4 P4 for $397!!  Yikes!
See after having to dump $20-$40 more for memory (RIMM's to DDR) and another $20-$50 for motherboard, your system is over $100 more for a system not as fast.

I used to buy nothing but intel, but just recently switched to AMD based for workstations.
Servers still get Pentiums, but that might change so as well.

Well, Thanks for your opinions... I ordered most of my stuff today...
xp2000
gigabyte ga-7vrxp
256 megs pc 2700 ram
Maxtor 80 gig 7200 rpm
SB Audigy
Lite-On 40x12x48 CDR
Enlight 7237 w 300 watt ps
sony floppy drive

only have vid card, speakers, dvd (wasn't in stock), monitor, and OS

edited to say: I couldn't justify extra expense for the P4... I'll just cover my case in pillows or something[:D]

I'm going geforce 4 ti4200 just need to pic a card and the features I want

I'll give you guys a [s]range[/s] basement report as soon as I get the stuff put together

It only makes sense when there is potential for a large increase.  Buying a 1.6 GHz northwood for $125 and running it at 2.4 (which normally costs several hundred) is significant.  I don't upgrade all that often, so it's nice to have a processor that I can ramp up later if I want.  Plus it's interesting [:D]

I wasn't smoking anything, but the folks at sharkyextreme must be:  [url]http://www.sharkyextreme.com/guides/WCPG/article.php/1151161[/url]

I estimated the prices from their "weekly CPU prices" guide, which is only 5 days old but is apparently completely wrong.  Thanks for the correction....  Next time I will spend the extra 5 seconds necessary to do my own search on pricewatch.

No, the DDR memory used is the same as that for AMD systems and Intel have viable DDR chipsets now.  Also the motherboards aren't any more expensive, so I don't know where you got that estimate.

MSI makes good motherboard for AMDs.  Its moderately priced and its very high quality.  As for the monitor, my CompUSA piece of xxxx 19" monitor is dying and I'm thinking about getting Sony Trinitron 17" or 19" display as a replacement.  Picture quality looks better on Trinitron and doesn't hurt my eyes as much.  I recommend ATI All-in-Wonder for graphics card, Plextor CD-R/CD-RW, and Western Digital harddrive.  I'm using several newer Maxtor harddrives but they also work fine but very noisy.

What part of "doing it right" or "Look at the Toms Hardware or overclockers websites" don't you understand.
Or perhaps you would like me to put up a warning post along the lines of the warning on the side of tobacco products.
Give me a break, I give the people on this board credit for having more common sense and intelligence than the average person.

Dude, are you in love with computer chips?  You sound like it is a very personal thing with you and chips.  You are the overclocker protector of computer chips or what.

"That your system has worked fine proves very little."

Yea, I guess it dosen't prove jack shit, it's just another bit of worthless "anecdotal evidence" not proved by actual use FOR THE PAST THREE YEARS.  I guess I shouldn't tell you that I was overclocking Celery chips (266-500's) with no problems at all.  

But in the end I guess I am just an electron fueled computer chip abuser.



"I have no problem with people overclocking if they understand the risks. It's your machine, you should be able to do what you want with it, and more power to you if you want to fiddle around with its insides. It's fun, if nothing else. However, I don't agree with encouraging people to overclock and telling them there are no risks.

That your system has worked fine proves very little. Integrated circuit reliability can be described statistically, and one data point, such as your system, proves basically nothing. When IC vendors gather reliability data on the operating life of their chips, they study 500 chips at a time. Generally, one failure out of 500 chips is acceptable, and two or more failures out of 500 is not. They raise the operating voltage by 10% and raise the ambient temperature to 125 degrees C, and this compresses 20 years of operating life down into one month of stress testing. No one has collected this kind of reliability data on overclocked chips. Overclockers rely on anecdotal evidence. But the scientific evidence shows that they are shortening the life of their chips. One thing that reliability testing in the industry has demonstrated is that some chips are weaker than others and will fail first while the majority continue to work. (I'm not talking about infant mortality failures here, I'm talking about end-of-life failures.) But eventually all of them will wear out due to gate oxide degradation, the hot electron effect, and electromigration."