Originally posted by: bryanW1995
they don't need it to be that fast since their $129 card is already faster than the 5870 in real world gameplay with lifelike physix effects.
lol wut? Blatant fanboi much? Not that I prefer Ati, I do prefer nvidia since my 6800 GT days on doom 3, and also because they fold better, but Cmon, this is just BS and makes no sense.
What $129 card are you talking about? Did you bring it back with your $129 billion dollar time machine? At present, you can get a good GTS 250 for $129, and MB a cheap GTX 260 for that on a good deal if you are lucky. W/E, the 5870 will probably be more than twice as fast as the latter.
Plus, even at $129 current gen, ATI wins in general in price-time per frame. {======}
The 4850 and sometimes the 4870 1 GB can be had for around that, with the first having an advantage in raw performance over nvidia's offering. GTX 260's may be better than the 4870 due to easier OC's, but generally cost a tad bit more. Still, a completely nonbiased person would usually be a wiser spender to go with Ati in this range unless they want to start folding.
Now, I personally prefer nvidia, as I fold and already have good experience with them, plus may want to get another 260. Or will wait for GT300. But I still try to be objective in my posts and considerations and can recognize when Ati would be the better choice.
////////////////////////////////////////////////////////////////////////////////////////////////
{- Please note that I use the price per performance ratio represented with the units of price-time per frame, using the logic that normally this would be measured as $/FPS = $/(frames/second) = ($ * seconds)/frame. Now technically a true ratio has no units, it basically a fraction.
So I guess price
erformance is technically not a ratio as they are measured with different units and :. at the end units remain as they cannot divide out.
If someone is of strong knowledge of scientific and mathematical reasoning and good with the use of units, dimensional analysis, AND preferably computer performance units a well, please correct me if my logic is flawed or I made a mistake or implemented my (theorem?) explantation wrong.
The way I see it, the "price per performance Ratio" is NOT a ratio. ( PPPerf != ratio) Given: monetary unit is different from performance unit...
<code>
#include <iostream>
#include "Ratio.h" //not going to be writing classes here though
#include "UnitPer.h"
int main()
{
UnitPer PPPerf;
PPPerf.numUnits = 3;
Ratio ratio;
ratio.numUnits = 0;
if(PPPerf.numUnits != ratio.numUnits)
cout << "QED";
}
</code>
OR
1. Ratio is a ratio iff it has no units, Ratio Definition
2. Price per performance "Ratio" must
be measured in "Monetary cost/performance unit", definition
3. Monetary cost unit is not equal to performance unit, given
4."Monetary cost unit / performance unit" has units of $/perf unit, division properties
5. :. "Ratio" has units, transitive property
"Ratio" != ratio, QED
-}
Facebook
Twitter