Damn, that's sort of confusing. Is that with GPUMark or GPUMark2 Ben?

 

 

Well, the file I downloaded was called 'GPUMark' so I presume it was GPUMark lol. - Would these be any possibilty to create a new thread with one post that includes, all directions, and the file for download? I think it would be a little bit less confusing because throughout the thread there are different instructions and different AA levels, downloads etc

 

Good post Ben

I would be interested to know what the difference is between my 560ti and the new 660ti before I spend my hard earned cash.......

 

Well, in these AA scenarios, the 660Ti has shown to be 21% faster than a GTX580. The GTX580 is around 30% faster than a GTX560Ti. So, if you upgrade to a 660Ti, your performance should improve by around 50% in heavy cloud + High AA situations.

 

Personally, for anyone who likes crisp quality without suffering performance, I would strongly reccomend that 660Ti. - This is easily the best bang for buck card I've ever purchased.

Thanks for sharing, Ben. Looks like you got a great card, at a great price.

Would these be any possibilty to create a new thread with one post that includes, all directions, and the file for download? I think it would be a little bit less confusing because throughout the thread there are different instructions and different AA levels, downloads etc

 

You're right Ben. Will post a new thread

Hey Ben it looks like my ATI 5850 is giving me better FPS than the 660TI. How exactly is that possible?

Hey Ben it looks like my ATI 5850 is giving me better FPS than the 660TI. How exactly is that possible?

 

ATI and Nvidia are not comparable in this situation. AMD does not have an option to force Sparse Grid Supersampling; If it did, I guarantee you that your frames would be much lower than the results of the 660Ti.

ATI and Nvidia are not comparable in this situation. AMD does not have an option to force Sparse Grid Supersampling; If it did, I guarantee you that your frames would be much lower than the results of the 660Ti.

 

Okay that sounds reasonable. But question. If what I see on my screen is smooth, jaggie free, and fluid, do I really care about Sparse Grid Supersampling? Don't I just want something that looks and performs great?

I have an idea Ben. Maybe we can both make a video with the same aircraft at the same airport taking off. Say we use the F1 Mustang, taking off from Bowerman airport (using Orbx scenery), and head north (I think) towards the mountains. Then we can have a better understanding of what we are each seeing. Maybe your settings make your experience much better than mine? I have no idea, and maybe I am missing something I don't even know about! And there is a slight chance my settings give me a pretty similar look to yours, without needing Sparse Grid Supersampling. My system would be the serious underdog, but I think it would be an interesting test.

 

BTW, just comparing specs you have 4.8GHz to my 4.0GHz, 2GB Video RAM to my 1GB, and a much newer video card. I also run 1920x1200 vs your 1080 resolution. And you have SSD's, and I don't.

 

So on paper this isn't even a contest.

 

You game?

without needing Sparse Grid Supersampling

 

Honestly, without SGSS, the GPU in FSX doesn't have much of a point.

It would be nice to see an example of it! lol

Good one! Looks like a fine card.

Now overclocked to 1137MHZ:

 

Frames Time (ms) Min Max Avg

7257 300000 21 29 24.19

Now overclocked to 1137MHZ:

 

Frames Time (ms) Min Max Avg

7257 300000 21 29 24.19

 

Keep going Ben! At some point you'll reach 5850 levels. Lol just kidding!

GTX660Ti Superclocked 3000MHZ Memory Clock:

Frames Time (ms) Min Max Avg

6878 300000 20 27 22.927

 

 

GTX660Ti Superclocked 3500MHZ Memory Clock:

Frames Time (ms) Min Max Avg

6720 300000 20 27 22.4

 

 

These findings are interesting... I guess you can interpret what this mean though haha

 

thanks for your findings :)

 

with reference to the memory overclock and benchmark results - it could be one of 2 things

 

(1) those results are well in the margin of error for any benchmark

 

(2) one thing to note DDR5 has built in error correction, which can start being used if overclocked too high. IE it will realise that errors are being made and will internally (on the memory itself) correct in real-time - BUT this does slow it down, therefore higher clocks can indeed get lower performance. Note that there are no outward signs (ie stability or corruption) - just lower performance.

 

I'm slightly confused why you're getting better performance from the 660Ti though as in most modern benchmarks its not really that much quicker than a 580, and it has much lower memory bandwidth than the 580 (which I'd have thought was useful for FSX)

 

personally I wouldn't upgrade my oc'd 480 for a 660Ti anyway

 

be interested in further results

 

I'm slightly confused why you're getting better performance from the 660Ti though as in most modern benchmarks its not really that much quicker than a 580

 

Even though it is only about 20-25% faster than the 580 in FSX, the smoothness is astounding...

My GTX285 has served me well for many years but now I'm looking to upgrade.

I'd like to know if the 660ti handles 8xS and 2xSGSS without any big frame drops in 1920x1200 (or 1080) and the heaviest weather possible.

And how about 4xSGSS with the 660ti, is that asking too much?

 

The 285 is almost capable of 8xS + 2xSGSS but falls just short.

 

In overall gaming a 285 -> 660ti would be a huge improvent, that's for sure.