Overclockers--are you fooling yourself???

[RickH]

Lets see--last month everyone loved AMD, this month it's the P4 1.6A. The AMD chips are fast, cheap, and pretty stable. The P4s are more stable, but even the P4 2.2 is not any faster in applications then the AMD PR 1.9, except you can say you have a 2+GHz machine. The problem is no one wants to spend the cash to get a "real" P4 2.0 or 2.2 so they overclock a slower chip. Even overclocked the P4 isn't much, if any faster than the AMD not overclocked. To make matters worse the overclocked P4 plateform is probably no more stable than the AMD and with the increased core voltage we are back to a hot running processor. Sorry, there is no free lunch. RRRRRRRRR

 

[z0ner]

<< Lets see--last month everyone loved AMD, this month it's the P4 1.6A. >>



Let's see... I don't think your 'flavor of the month' analagy is quite correct. AMD has been an overwhelming favorite in the O/C community for at least the last 18 months.



<< The AMD chips are fast, cheap, and pretty stable. The P4s are more stable, >>



Yeah, stability's a bitch, aint it?



<< but even the P4 2.2 is not any faster in applications then the AMD PR 1.9, except you can say you have a 2+GHz machine. >>



I agree, in most applications that aren't FPU intensive the Athlon XP 1.9GHz is on par with a P4 2.0 GHz. But,

1) most people are getting 2.3GHz out of their 1.6GHz Northwoods, which kills AMD in ANY application
2) the P4 is optimized for FPS games, so even an equivalent Athlon would be slower
3) with the P4 you're getting rock solid stability
4) 2.3GHz in most cases with the QUIET RETAIL FAN. Thanks to Intel I'm starting to get my hearing back!
5) oh yeah, and I'm saying that I have a 2.3GHz machine 44C under full load with retail stock fan.



<< Sorry, there is no free lunch. RRRRRRRRR >>



You really need to educate yourself a bit better on this topic, and embrace the better technology.

 

[Theslowone]

Are you making an arguement or a statement? People overclock because they want to and because it is cheaper. Right now the p4a's are doing pretty darn well and while still more in cost they seem to be fun to work with and perform very well. yes the Athlon still performs better in some instances but when you can get a p4 up to 2.5 ghz then thats doing something. I own a athlon, and happy to have one, but your arguement sounds like a rant, and its full of holes. And there is a free lunch most of the 1.6a and 1.8a are very stable with retail hs/f at or above the 2ghz mark, so how is this not free? Getting the clock speed of a cpu that isn't even out yet for the price of a 1.8a P4. Take off the blindfold and look at both sides of the board.

 

[Duvie]

I agree with some of your thoughts....

I think your analogy of a 1.9 amd equaling a 2.2...maybe a willamette core, but not a northwood core...a p4 2.0a is equal to a 2000+ xp...Now my 1.8a at 2.4 will likely beat any amd unclocked out there...probably equal to a 2200+ xp...

I am going to get pc2700 ddr to get the higher bandwidth and that will helptrmendously...

I even have a bit left...I am only at 1.57v (limits of the crappy msi board)...temps are a bit high so I have ordered a better fan and AS3...I will try 2.5 and above...

I have ran 2.4 at 1.57v for 8 hours plus with no errors in prime 95...

I get errors at 2.52 right out of the gate but I am confident it is voltage nbeing delivered (along with jumping msi voltage) and slight heat...

 

[RickH]

<< Let's see... I don't think your 'flavor of the month' analagy is quite correct. AMD has been an overwhelming favorite in the O/C community for at least the last 18 months. >>



[A] Are you sure it's 18 months, not 15 or 20. The number value is not to be exact.



<< I agree, in most applications that aren't FPU intensive the Athlon XP 1.9GHz is on par with a P4 2.0 GHz. But, >>



1) most people are getting 2.3GHz out of their 1.6GHz Northwoods, which kills AMD in ANY application
2) the P4 is optimized for FPS games, so even an equivalent Athlon would be slower
3) with the P4 you're getting rock solid stability
4) 2.3GHz in most cases with the QUIET RETAIL FAN. Thanks to Intel I'm starting to get my hearing back!
5) oh yeah, and I'm saying that I have a 2.3GHz machine 44C under full load with retail stock fan.

[A] Check out this on THG, you probably think Tom's is worthless, but they have been at this a long time.

http://www6.tomshardware.com/cpu/02q1/020128/index.html



<< You really need to educate yourself a bit better on this topic, and embrace the better technology >>



[A] I have been at this a long time also. The old Celeron 300A, 533A, etc spoiled us because you could overclock them and still keep everything in your system with normal operating conditions. I just don't want people who know very little about computer hardware to spend hundreds or thousands of dollars on a overclocked system that is a pain in the butt.

 

[MilkPowderR]

RickH is just whining. Stop your bitchin'. hahaha. Tap tap tap!!
To be honest, I'm not that crazy about Woodie P4. Their performance seem still too weak for its "mhz" raw power. It is the fact guys.. Don't hide the truth. RDRAM P4 systems are different story... maybe an exception. My gut feeling prediction tells me that the Throughbred Athlon will outperform P4 DDR's no matter how far they're OC'ed. A descent XP DDR system OC'ed to a descent level does 9000+mark on the 3Dmark 2001 with a GF3/R8500. I have seen many of P4 DDR systems OC'ed 2.4Ghz- 2.6Ghz doing 8500- 10,000+ mark on a GF3/R8500. RDRAM OC'ed systems are still being observed by me. I don't have much words to say about that yet. So, way I look at it, P4 sucks. Too weak for its mhz.

 

[MadRat]

<<I agree, in most applications that aren't FPU intensive the Athlon XP 1.9GHz is on par with a P4 2.0 GHz.>>

More like on all non-optimized benchmarks.

<<But, 1) most people are getting 2.3GHz out of their 1.6GHz Northwoods, which kills AMD in ANY application>>

It does? As far as the reviews have shown it is up to par, not exactly killing AMD. Heck, it was a pretty dead heat when they compared the 1900+ XP to the 2.2GHz P4; 600MHz advantage all for not.

<<2) the P4 is optimized for FPS games, so even an equivalent Athlon would be slower>>

The P4 is what?? The Pentium 4 is not optimized for more FPS games than for what it is optimized. The Athlon/XP chips perform more healthy than the P4 on the Quake2 engine games.

<<3) with the P4 you're getting rock solid stability>>

With overclocked P4's this is not the case. People are reporting alot of psuedo-stability as rock solid stability.

<<4) 2.3GHz in most cases with the QUIET RETAIL FAN. Thanks to Intel I'm starting to get my hearing back!>>

The AMD retail fan is quiet, too. Alot of 5000rpm fans can keep the AMD XP-core cool at a true 1.67GHz. As soon as Thoroughbred is released you'll see people touting how quiet are the retail fans.

<<5) oh yeah, and I'm saying that I have a 2.3GHz machine 44C under full load with retail stock fan.>>

Temps mean nothing. Stability has more to do with the workings of the inner core than temperature. Pushing the 1.6a P4 to 2.3GHz (especially using DDR RAM) is not any better than pushing a 1500+ XP to a 2000+ XP-equivalent. Using DDR RAM with the 2.3GHz P4 is almost as silly as pushing the old celeron 266 (non-A) to 500MHz.

 

[RickH]

Don't get me wrong--I want you to be right. I am tired of AMD heat and sorry ass chipsets. I have had the money for a new machine burning a hole in my pocket for months. Just be careful in the forms--danger is lurking--your dream machine could turn out to be a nightmare!!!! There are many people with thousands of posts to their credit giving wrong information. Second hand information heard somewhere maybe fine for you, but if you are advising people how to spend their money--please make sure you are correct in your statements. R

 

[Daovonnaex]

3.2GHz...I doubt that I'm fooling myself. High voltage isn't a concern, as I replace CPUs frequently.

 

[RickH]

<< 3.2GHz...I doubt that I'm fooling myself. High voltage isn't a concern, as I replace CPUs frequently >>



Well anyone that would build a machine that could fail at any instant due to thermal, mechanical or electrical problems is a fool.

 

[Yourself]

<<

<< 3.2GHz...I doubt that I'm fooling myself. High voltage isn't a concern, as I replace CPUs frequently >>



Well anyone that would build a machine that could fail at any instant due to thermal, mechanical or electrical problems is a fool. >>




I guess we are all fools then....that can happen on any PC built, whether it's run in spec or not.


Self

 

[Tom]

Madrat,
"Heck, it was a pretty dead heat when they compared the 1900+ XP to the 2.2GHz P4.."
reply- That was a 2.2g P4 running at 100fsb which is considerably slower than a 2.4g P4 running at 133fsb(which is what you get with a 1.8a oc'd to 2.4)

"With overclocked P4's this is not the case. People are reporting alot of psuedo-stability as rock solid stability."
and "Stability has more to do with the workings of the inner core than temperature. Pushing the 1.6a P4 to 2.3GHz (especially using DDR RAM) is not any better than pushing a 1500+ XP to a 2000+ XP-equivalent"
reply- running a 1.6a or 1.8a Northwood at 133fsb is well within the design limits of the chip. And most of the "Stability issues" of AMD products are because of the chipsets that most AMD users have, not the AMD cpu. With an Intel cpu the user has the option of using an Intel chipset, which have the best reputation for stability.

"The AMD retail fan is quiet, too.."
Not even close to as quiet as the Northwood retail HSF. Additionally the Northwood HSF is all the cooling needed for an overclock to 133fsb and even farther.



RickH,
The Toms Hardware article points out several things that actually argue in favor of a DDR motherboard and an overclocked Northwood. An rdram based system is even faster, but my opinion is that for an rdram based system it would be better to wait a couple of months for the 533fsb rdram boards. In the meantime the DDR boards, can be the basis of a very economical system that is faster than almost any non-overclocked system, more stable than other overclocked systems, Amd or Intel, that run out of spec FSB speeds, which adds up to a very fast and stable system for very little money.

And I agree with your comments about overclocking, it's not as simple as some would make it sound. There are lots of potential problems when building any computer, overclocking just magnifies this. But if playing with computer stuff is something one enjoys then these new lower clocked Northwoods are interesting.

 

[Theslowone]

So what is a stable system that will not fail becuase it is so good? For that matter what machine is, not just computers but spacecraft, aircraft, cars, boats. Everything we have from electric razor to nuclear engines can fail just like that from any of the problems you just listed.

 

[Duvie]

A bunch of you still have numbers all messed up....

A p4 at base 400fsb the ddr will not suffer that much compared to the rdram...

However when you have the ability to couple a p4 northwood with a 333ddr board like the sis chipset you can run pc2700 and pc3000 equivalent memory why cranking fsb up to 533, which many rdram boards can't do yetand reviews have shown the sis chipset on a 333dr platform is comparable....around 95 percent....

I love athlons still own one, but don't fool yourself if you think you will be able to run with my board 1.8a @2.4 with 533fsb and 333mhz memory with an athlon 1900, or a 2000+ xp oc'd to 2200+....

I don't play games...I run multimedia appz and cadd...I know the athlon was better at the cadd but I have jumped enough up to be still in range of a higher athlon processor in performance...In multimedia appz I will be faster...ie like mpeg4 and divx encodings...



I have also noticed the silence of this machine...I was able to remove one of my case fans and the HSF runs more silent then my athlon hsf...


Also to the ppl making up the numbers a recent article comared the sis, 845, and i850 chipset and at default (no ocing) i845 ddr platforms were about 90% of rdram...so quit making up numbers....

My i845 mobo at hgher fsb locks my ram down so as I can never get higher then 266mhz or pc2100...with my sis board coming I can clock the memory higher and get much more bandwidth.

 

[manko]

Well, the reasons I just ordered a 1.6a are:

1. I couldn't wait for the .13 micron Thorougbreds (but I would have, if these damn 1.6as didn't OC so well)

2. AMD pissed me off by not letting me win a free XP1800 kit at their giveaway, despite waiting out in the cold for 4 hours

3. it's cool and quiet with a bigger cache (than old p4s)

4. the new boards have modern features like usb2 and ata133, and can handle ddr333 (again couldn't wait for Thoroughbred mobos)

5. I picked up the $99 ti200 and it really sucks in my Celery300a@450 machine

6. the thought of paying $140 for a $500+ proc (because AMD doesn't ream customers as much at the high end, I guess)

7. I wanted to avoid all the potential VIA headaches (I have an old VIA board in one of my machines...nothing but trouble)

 

[DAPUNISHER]

Despite RickH's less than tactful approach, the cause for his irritation is justifiable.

<< There are many people with thousands of posts to their credit giving wrong information. Second hand information heard somewhere maybe fine for you, but if you are advising people how to spend their money--please make sure you are correct in your statements. R >>

Agreed and seconded. I have a 1.6A/P4B333/2700DDR combo arriving by wednesday and my reason for ordering it is to ascertain if indeed it is worthy of the hoopla(I admit I'm somewhat sucked in by it as well ) . I just sold an XP1500+/K7S5A/PC133 SDRAM system for 50$ more than I'm reinvesting in the P4 Rig so I don't foresee an inequitable experience on the horizon. I also have not built a Intel based system in 2+years being an AMD lovin' kinda dude so it'll be a (hopefully pleasant) change of pace. However, I'm under no illusions that it'll be anything but an ephemeral experience. When the Thoroughbred arrives it is my belief that it'll ensure AMD's continued success as the leader in high performance processors. As to the AMD vs Intel theme which tends to dominate such threads, I find it exasperating and lamentable From my perspective it's all good, I don't hold stock in either company, I'm not directly employed by either and both are making stable, reliable processors and platforms that provide exceptable computing solutions and help ensure competitive pricing will continue. Hence I'm both optimistic and encouraged by the heated competition that has developed in the last several years........all I need is more popcorn cause the show is gettin' good

 

[Theslowone]

There are many people with thousands of posts to their credit giving wrong information. Second hand information heard somewhere maybe fine for you, but if you are advising people how to spend their money--please make sure you are correct in your statements. R

What other info do you have that is not second hand or our own experience, people give free advice to people on what they recommend. Their suggestions aren't written in stone. Some people enjoy overclocking, and they are gettiing great results with the new p4s, so whats wrong with this? They recommend what they would get or what they got. This board is far from all facts and people coming here are looking for opinions. If they werent they wouldn't ask what do you think of this system. Or what should I get. Just because you don't believe somethings not right doesn't mean your right. And I will admit that I have been wrong at least once . Post count doesn't mean your a genius or that you know the whole truth behind the components involved, just that you have posted a lot.

 

[DAPUNISHER]

Hey Theslowone
I interpreted that statement in a different context than your applying to it. I saw it as an admonishment to those who recommend hardware they have not personally evaluated and who are often considered credible sources of information by many newer members of the forums based on their member ranking/years of membership here. As to your being wrong once....say it isn't so!!!:Q

 

[jiffylube1024]

I agree with the comments about "pseudostability as rock stability" as stated above. For many picky people like me, it has to be so rock stable that one single crash/error is too much. I'm running a 1.8A @ 2.25 GHz right now... that IS stable as a rock for me. 2.4 just doesn't meet my personal standard of stability (maybe I have a bad chip or maybe I'm just picky - it seems pretty sweet to me). Btw, the P4B266 runs 125MHz FSB with a 31 MHz PCI, so its actually underclocking it a tad.