As a result of this we were able to find a fairly accurate regression line to the data. BioShock Infinite and Metro: Test system and drivers Testbed: Consumed Power, Energy Consumption: Anyway, some cards in the overclocked mode came close to the Ti For the comparative analyses we used the following cards:

Uploader: Shakasida
Date Added: 12 January 2012
File Size: 28.45 Mb
Operating Systems: Windows NT/2000/XP/2003/2003/7/8/10 MacOS 10/X
Downloads: 67833
Price: Free* [*Free Regsitration Required]

Does this mean that you should only worry about overclocking your card’s core? The most splendid package!

NVIDIA GeForce2/3 Titanium Roundup – January 2002

That is why it was surprising to see that the 5ns worked stably at MHz. Using this line we can see that each 1 MHz that the core clock is pushed, our game performance increased.

Pentium 4 based computer Socket The Gainward’s card has the 4. Overclocking only the GeForce3 Ti core provided a noticeable speed increase. Of course we can calculate what memory is needed for the toughest 3DMark test in x, but do you need those fps gained?

Much like we saw with the core overclock, performance did not level off as the memory clock speed got higher, suggesting that the memory bus does act as a bottleneck of some sort.

Consumed Power, Energy Consumption: This could be due to the fact that we were forced to stop overclocking the memory bus of the card at MHz so we therefore used this memory speed at the core speeds of MHz and above.


Sandy Bridge Trying out the new method. Production of such models is just a marketing trick taking into account that the whole GeForce3 family is going to die soon the GeForce4 Ti line will press them out quicker than games are going to need MBytes of memory. Just at the beginning geforc3e the year we discussed whether it was good or not for video cards to have MBytes memory.

Overclocking the GeForce3 Ti – NVIDIA GeForce2/3 Titanium Roundup – January

Using gianward information it seems that overclocking the core clock of the GeForce3 Ti provides more bang for the buck than overclocking the memory bus does. The tests revealed that such memory size is not needed yet.

As a result of this we were able to find a fairly accurate regression line to the data. By far this is a much larger speed increase than we saw when we overclocked just the core or just the memory alone.

For testing we used the GeForce3 Ti card that we were able to overclock the highest: The peculiarities of the Leadtek card were considered in detail in the review of MBytes cards. I already mentioned it examining the card with 64 MBytes memory.

Although, considering that the Ti cards are equipped with the chips which don’t not work stably at MHz and, thus, do not suit Tithere is nothing to be surprised at. Unfortunately, there is no heat-conducting grease between the heatsink and the memory. But now the designers decided to make a flower of it.


The heatsink is attached to the card the same way as on the Leadtek solution: Increasing the core speed from the stock speed to the maximum core speed possible resulted in a Write a comment below. The slope of 0. We have proven it in that review, and our 3Digest shows that there are no games able to enable the memory beyond 64 MBytes.

Surprisingly, there are no trails of cupeling and a label with a seal: The Second Encounter Grand Cathedral, quality mode The tests were carried out in a bit color mode.

ShieldSquare reCAPTCHA Page

For the comparative analyses we used the following cards: All the cards ship in retail packages apart from the Palit’s solution which is available both in a retail box and in an OEM package, though the accessories are the same. I like it very much.

The trend in the data gaainward when overclocking the core and the memory of the GeForce3 Ti seemed to decrease gradually as clock speed increased. For those of you with a bit more math knowledge, you may notice that the data above fits a logistic regression line better than a linear regression line.