View Full Version : What is better: OC or no OC?
Mr.Radar
02-10-04, 07:36 PM
My sig system has the crappiest RAM on earth. It won't go too much over 166 w/o failing prime, so I set the FSB ratio to 6/5. With that setting, I am able to get 3200+ speeds without failing prime (I have yet to test beyond 9 hours though). I was wondering, what would give me better performance? Unoverclocked with 1/1 or overclocked with 6/5? I can't run any benchmarks on my system besides PCMark '02 because they all require an AGP video card. I think I might have one around here somewhere, but I don't want to open up my system unless I really have to.
Edit: Right now I'm going to run PCMark and compare the scores.
Mr.Radar
02-10-04, 08:01 PM
Here are the PCMark 2002 results:
CPU: This one was no surprise. 6684 @ 3200+, 5511 @ Stock. A 1173 point improvement.
HD: 1297 @ 3200, 1298 @ Stock, nothing to report here
Now this is where it gets strange. By common knowledge, I should get a marginally worse score on the memory test, however...
Mem: 4692 @ 3200, 4312 @ Stock. A 380 point improvement.
This is just one artificial benchmark, but the scores seem to indicate that both CPU AND Memory are faster even with the wierd FSB ratio set. Before I start P95'ing at Prio. 10, I'd like to hear some outsize opinions on this.
Why such a crappy video card on this system?
Mr.Radar
02-10-04, 08:20 PM
Originally posted by Fleck
Why such a crappy video card on this system?
A) It's been a good card. The AGP card I referred to above is worse than a GeForce 2MX. It was a Mad Dog Multimedia Prowler64 (or something) I saw it for $50 on sale at CompUSA a few years ago (before I knew anything about video cards) and though "Ooooh 64 Megs! And it's 2x AGP!" so I got it. (I paid the same price for my AiW card a year before so I was expecting better). That card was going to replace my current card in my previous system (which was nothing but a disaster in itself (1.2GHz Duron in a KT133 board)). Even the demos the card came with ran very choppy and Half-Life ran very crappily on it, even in software mode (which I still find unbelieveable). It's drivers never worked right and the TV-Out was set to PAL mode and the only way to change it to NTSC was to change a setting on the TV utility that I could never get to install. It had an in-house developed GPU on it and went out of production about a week after I bought it, which tells you something. Man I've gone on about this too long! I guess that the most important thing I learned from this card is to research any products you buy through commercial reviews & customer reviews.
B) Video capture is my second use for this system (besides internet & music) and this card does a good job at it.
C) I can't afford a new card.
My video card is first on my list of things to upgrade. [Right now I'm looking at the Asus 9600XT/TVD card, but I'm going to wait until the RV40 cards come out so the price will drop on it.]
dead_man311
02-10-04, 08:41 PM
my sister has the same video card and i thought for a lower end card it work really good. she plays alot of DVD's and play some smaller games alot and the card works perfect
Mr.Radar
02-10-04, 08:45 PM
The ultimate irony is that if I get the 9600XT I won't be able to use it with a BTX machine (no AGP slot) but I'll still be able to use my Rage128 Pro because it's a PCI card. Now, not to be rude, but could we get back on topic? I'd really like to know if I should prime it or not before I spend 24 hours on this.
Edit: I think I'm going to start primimg it anyways.
barton2500
02-10-04, 09:33 PM
9 hours is plenty in my opinion. It's possible to have unstability, but that only applies if you actuall use the CPU at 100% load for over 9 hours. Which typically doesn't happen unless you are rendering or encoding.
You could prime it overnight so you can use it during the day.
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.