Which is important for a better performance of a game? Pixel Rate or Texture Rate?

Question

I've no idea of buying any of these products as they are out of my budget.. :p But just out of curiosity, I wanna know which one is better.

http://www.game-debate.com/gpu/index.php?gid=1583&gid2=1105&compare=radeon-hd-7950-sapphire-oc-3gb-edition-vs-geforce-gtx-660-ti-zotac-amp-edition

According to this website, the specs of hd 7950 are much better when compared to that of the GTX 660 ti. (Like Memory, Memory Bus, Memory Bandwidth & Pixel Rate). But the Texture Rate & Memory Speed of GTX 660 Ti AMP are somewhat better. But what I felt was, I mean after looking at those specs, HD 7950 is way better than GTX 660 Ti AMP.

But the final result was that GTX 660 Ti AMP was better than HD 7950. I wanna know why?

And which is important for a better performance? The Pixel Rate or the Texture Rate?

Answer 1

You just discovered the big truth about graphics cards.



You cannot extrapolate real-world performance from numerical specs. The biggest factor in how cards perform isn't core or memory clock speed, memory bandwidth, amount of VRAM or anything represented by a numerical value. It's the GPU core a card is based upon. Just like knowing whether a CPU is a 2nd-generation Core i5, Phenom II, A-series, Bulldozer FX, Piledriver FX or 3rd-generation Core i7 is more important than the cache size, clock speed etc.



You really can't break down why Intel processors are better than equally (or even higher) clocked AMD processors- Intel's Core architecture is simply more efficient and powerful. so a first generation Core i5 760 running at 2.8Ghz outperforms a 3.4Ghz Phenom II X4 965 running at 3.4Ghz. Simply comparing spec values doesn't translate- you have to check benchmarks and lab tests to see which CPUs actually perform better. It's the same thing with GPUs. Some GPU designs are more powerful than others. Secondary specs don't change that, and manufacturers constantly tweak those- some combinations work well, others don't.



For example the GeForce 9800GT was a 256-bit GDDR3 card. But it's slightly slower than a Radeon HD 4770 which was a 128-bit GDDR5 card. It turned out the smaller memory pipeline leading to faster memory worked just as well as the larger pipeline to slower memory, at least at the popular resolutions of the day (1440x900 and 1680x1050). Yet the Radeon HD 4770 was still slower than a GTS 250 (256-bit GDDR3 card) which was slower than the 128-bit GDDR5 Radeon HD 5770.



What matters isn't which cards have more processor cores or higher clock rates, but how it all works together in real-world gaming. Graphics cards are judged by performance, not by spec.



Between the stock GeForce GTX 660 Ti and Radeon HD 7950, the 7950 is slightly faster in most titles. The GTX 660 Ti AMP version is about a draw overall. However some games are coded so they favor AMD's architecture (Skyrim, Arma II, STALKER Call of Pripyat) while others favor Nvidia's architecture (Battlefield 3, Borderlands 2, Total War Shogun 2 and especially Starcraft 2). So you'll want to check performance in the titles you play most. Also, note that the GTX 660 Ti seems to be optimized for 1680x1050 resolution while the HD 7950 performs a bit better at 1920x1200 (having a larger memory interface tends to help at resolutions of 1920x1080 and higher)



http://www.techpowerup.com/reviews/AMD/Catalyst_12.11_Performance/5.html

http://www.tomshardware.com/reviews/tahiti-le-7870-7930-benchmark,3401-3.html

Answer 2

This Site Might Help You.



RE:

Which is important for a better performance of a game? Pixel Rate or Texture Rate?

I've no idea of buying any of these products as they are out of my budget.. :p But just out of curiosity, I wanna know which one is...

Answer 3

Pixel rate for the high Res but most of us are using 1980x1080 res and both will work in that res most important for gamer is the texture rate = the quality of the game you need both actually but texture>pixel for gamer

Answer 4

For the best answers, search on this site https://shorturl.im/avc4J



hi mate out of all these i would say cuda cores ( also known as shader processors) ,however the graphics clock is not a official specification.... theres the core clock which would in theory be a processor clock ( gpu clock or graphics processing unit ) ..... and this is one of the most important factors when it comes to video cards..... and then theres the memory clock which along with the core clock are in my opinion the most important factors when choosing a video card,especially for gaming the texture fill rate is a minor specification,and along with pixel fill rate and flops(floating line operations ) are the speed at which textures and pixels can be drawn/processed in every second ( normally in the billions ) another part which is commonly overlooked is the shader model ( also known as pixel and shader model ) and this is allways combined with the same direct x compliance... so shader model 4.1 is allways combined with direct x 10... and shader model 3.0 with direct x 9.0 ....... and you will also find direct x and shader model will also be associated with the same open gL version... and with direct x 11 and shader model 5.0 you will normally find open gL 4.1 when dealing with pre direct x 10 cards they didnt use shader processors,instead there were pixel pipelines .. and theres were inferior to shader processors ( or cuda cores) .... pixel pipelines could only work with pixels and then geometry pipelines and texture pipelines were needed for geometry and texture calculations however shader processors and cuda cores can do pixel,texture and geometry calculations on the same processor ... this saves time and is quicker since previously if all pixel pipelines were full the process would need to wait until one was free,however since shader processors can do all graphic processing the calcuation or process can go to any of the free shader processors... and the more shader processors or cuda cores you have the better the performance. note : nvidia have less cuda cores than ati,s equivelent shader processors... however this is because ati/amd work their processing cores out differently to nvidia.... as a rule there are 4 ati shader processors to every 1 nvidia cuda core ... in a nutshell the most important factors in a graphic card for gaming are 1) core clock ( speed at which the gpu operates at ) 2) memory clock ( speed at which the cards memory runs at ... eg 2000mhz ..... however the effective rate is calculated by the type of memory used... eg 1200mhz standard memory speed.... x gddr5 memory ( quad pumped ) = effective speed of 4800mhz 3) memory bandwidth ( the speed at which the card can access memory. ) memory bandwidth is equal to the size of the memory bus multiplied by the speed at which the memory is clocked.the higher the memory bandwidth, the better the card will be able to handle large textures and anti-aliasing and anisotropic filtering 4) memory interface/bus width ( the number of bits wide (and the organization) of the memory bus that connects your GPU to your video card RAM. ) these are in my opinion the main factors to consider when buying a card for gaming.... if these are good then generally the others like the texture fill rate will also be good..... i hope this helps,any questions let me know good luck mate and happy xmas !