Been reading slizone and was surprised to see that the officials of SLI Zone claimed that it's a waste of money paying extra for factory pre-oc cards.

As we all know, alot of manufactures tries to stand out from each other when it comes to selling their "value-for-money" products. Increased clocks, redesign cooler (and even partnership with renown cooling specialists), longer warranty, 90-days step up program, gold plated DVI connectors are some of the examples the manufactures do to outshine their competitors.

As most of us are interested in performance rather than anything else, do you think it's worth paying a little extra for the pre-oc cards, or, rather buy the reference speed cards and OC it yourself?

does preclocked cards can oc better than default clock cards?

dun think it applies only to pre-oc cards. i think this approach is to make themselves unique from their competitors.
asus 8800gts 320mb for example, is a reference card, but it uses 1.1ns memory (forgot wat memory brand) as opposed to 1.2ns.

are you sure there's a variant of 79gs at 600/800 from galaxy? AFAIK, the highest 79gs sold by Galaxy is 560mhz (the one that comes with zalman fan).

you are right abt the 1.2ns RAM, other 79gs uses 1.4ns. i'm using 1 now

As tweaking program like RT and ATI Tools can help to overclock an OEM cards to (or probably higher) pre-oc cards with stability, there's 1 thing that we cant OC; shader clocks.

Here's what I've noticed from both my 8 series gc. my reference card (Asus 8800GTS 320) is running at about 1188mhz shader clock. As shader clock increase in propotion to core clock, I've managed to have a stable OC with 1418 mhz shader clock (from stock core 500mhz to 631mhz).

On the other hand, XFX 8800GTS XXX has a pre OC core clock at 580mhz (reference clock 500mhz). WHen I OC the XFX's core to 631mhz, the shader clock jumps from 1512mhz to 1620mhz.

Here's a summary table of both the cards;

Asus 8800GTS 320 MB | 500/1600/1188 (core/memory/shader)
XFX 8800GTS 320 MBXXX | 580/1800/1512 (core/memory/shader)


As shader clocks increases when core increases, I've set the core clock to 631mhz and compare both the shaders.

Asus 8800GTS 320 MB | 631/1884/1418 (core/memory/shader)
XFX 8800GTS 320 MBXXX | 631/1884/1620 (core/memory/shader)

The result shows that pre-OC cards hv higher shader clocks compared to reference card despite the same core clocks.

As of now, neither RT nor Ati Tools (pls correct me if i'm wrong) is unable to control the shader clock, Getting pre-oc cards seems to be the better deal here; as far as the unified shader domain is concerned.

Just my 2 cents.