CPU speed VS. stability

[fixrupr]

What relationship is there (if any) between CPU speed and system stability?

I have Athlon 1200 266fsb system, as well as a couple of Duron 800 machines. I see some difference in performance, but not a great deal, while running real world apps (office progs no games). My programs run acceptably on an AMD 400 w/ 128mb.

Am designing new system (non gamer) and find that fast athlons are dirt cheap (1400 266fsb @ $99 or Athlon XP 1600 @ $123). But, does that speed have some baggage with it?

I can’t make heads or tails out of benchmark testing graphs. Tests that deal in nano seconds might matter to folks running Quake but not to office machines where reliability is the primary concern.

I might mistakenly be using knowledge of auto engines to question if there is a tradeoff between stability and speed. A 2.5 litre 400hp racing machine will not haul groceries reliably. Our computers “haul groceries”, but I want the trip to the store to be safe first, and fast if possible.

Sorry if I’m rambling a bit here, the bottom line question is:
Is there a point of diminishing returns re CPU speed on an office machine? (stability the issue not $$)

Some of you guys work on office machines for your bread and butter, and therefore know what is best. I wish to sip from your font of knowledge, if I may.

------------------------------------------------------------------

“I have offended God and mankind because my work did not reach the quality it should have.”
------- Leonardo da Vinci-------

[Randy_tx]

Although the "Benchmark" tests used by Manufacturers to access these things do, in fact, demonstrate the newer CPU's are indeed faster....most of us could never tell it after about 800 mhz! There is one advantage in buying the newer higher rated CPU....they may actually be more stable than the earlier models [esp in the AMD lineup]. AMD has learned a great deal about how to lower cpu heat for example and the newer chips do run cooler....thus more stable. Additionaly, AMD has made a number of other improvements in the newest offering [ the XP series of cpu's]which you can learn more about at http://www6.tomshardware.com/cpu/01q4/011105/index.html.


------------------
Will XP save Me ?

[andreadebiase]

Originally posted by Randy_tx:
Additionaly, AMD has made a number of other improvements in the newest offering [ the XP series of cpu's]which you can learn more about at http://www6.tomshardware.com/cpu/01q4/011105/index.html.



so, from what you are saying I infer that if you have, for example, a Duron 700, it is smarter, for stability purposes, to spend more money and change mobo that accepts an Athlon XP (cooler and more stable) rather than to keep the same "old" mobo and push it to the maximum with an Athlon Thunderbird 1200MHz? did I get it right?

------------------
#!@$*%.....just format the damn C:

[Randy_tx]

This stability thing is "relative"! I'm running an old "Slot A" 900 mhz Athalon and although I do occasionally have to reboot and I do have to pay attention to the cpu temps......it runs just fine. The BIG improvement on cooler cpu running temps came with the XP series of cpu's for AMD. They went to the 13 micron wafer which somehow runs cooler....maybe a wafer guru could tell us why!

------------------
Will XP save Me ?

[fixrupr]

Thanks for your reply,
tomshardware.com is very good , and a prior visit there is what prompted my original inquiry. All benchmark graphs show P4 and fast Athlons leading the pack. Are not all these benchmarks purely speed related? And, running Win98se and / or W2k, which benchmarks are the most important to me?

[Randy_tx]

Benchmarks are unimportant unless you are a VERY serious gamer! Most of the business/personal finance software out there cant really take advantage of much faster than 700 mhz cpu[with 128+mb Ram] from what I've seen. If you have broadband access to the net and want to play games on the net...Get the fastest Athalon XP processor available and you will be "King of the Hill"! For all other computing, I never pay attention to Benchmarks because I know the software isn't going fast enough to care whether it's a PII, PIII, P4 , Athalon or Athalon XP! All of them are good for business/personal software.

------------------
Will XP save Me ?

[fixrupr]

Thanks again Randy,
From what I hear you saying, CPU heat is the primary concern if one were to install a fast processor. Am I to assume from your information that an Athlon XP 1500 would tend to be more stable than a Duron 1000?

-----------------------------------------------------------------------
"When choosing between two evils, I always like to try the one I've never tried before. "
---- Mae West (1892-1980)----

[ranchdog]

The thinner the wafer the less power draw. Less power draw to cause the wafer "vibrate." Less heat generated.

The Duron 1K is made on the new Morgan core so it's now also a cooler running chip. The Morgan core hit the scene with the XP's and Palamino chips.

Stability and office/workstation situation.... I'd choose a solid Mboard that had a reputation for being stable before I done anything. Then get myself a proccessor that would give me the most bang for the buck. Right now it is the XP 1700 Series.

Luck.


http://www.PCGuide.com/ubb/wink.gif

------------------
....How long is a minute... depends on which side of the Bathroom door you're on. ......Indecision may or may not be my problem......
..........

[This message has been edited by ranchdog (edited 12-10-2001).]

[fixrupr]

Ranchdog,
Thanks, your information is valued.
I’ll do some more research now, based upon the facts as presented. At present I’m leaning towards the Duron 1000, mainly because it might be best to wait a year till the dust settles around the XP scene. In the meantime, have you a recommendation regarding RAID motherboard with “solid reputation”? Combo of standard mobo and raid controller card would also be considered.
Opinions of those possessing experience in non-gaming systems are welcome.
thanks all! http://www.PCGuide.com/ubb/smile.gif
------------------------
PS actually found a site that rates stability in review testing.
http://www.motherboards.org



[This message has been edited by fixrupr (edited 12-11-2001).]