GTX580 vs GTX680 comparison

[Noel 4,563]

Huh? My Corsair 400R has 2 120MM front intakes, a bottom 140MM intake, 2 side 140MM intakes, and 3 120MM exhaust..... My case has plenty of case ventilation.

 

Also, you have to remember that there are many factors to temps, different ambient temps, different GPUs, and GPU voltage. I wouldn't ever stereotype an entire brand as bad just because one product might be flawed.... Corsair in general has some awesome products and so does Antec. They both also have there bad cases...

 

I have run FSX w/ the side cover off of my Lian Li case for many years now. A thru-the-wall A/C unit blows either room air (during wintertime here) or cooled air (during the hot months of summer here), at a low flow rate to reduce/eliminate static build up, over all components in the case. I don't have problems w/ dust or anything. Not sure why more folks don't take advantage of this approach. It can be a little noisy, but since I often use headphones this doesn't come into play much. I have cool running main board, ram, CPU. During summer w/ an over clock of 3.8Ghz on my old Q9650 I get temps of only 50C or so.

[vyper883 20]

I will report back late this evening for sure.

 

Hi guys:)

 

Sorry for taking so long to actually get back to reporting on the 480 vs the 680 in my i7 920 rig. I was busy building a new i7 2700K system, along with win 7 tuning/tweaking/fresh FSX install. I am REALLY Pleased with the performance of the Sandy bridge, and I have not even overclocked yet! However I will leave that aside for now.

 

Well, as I said I was not expecting a quantum leap in performance regarding 480 vs 680 EVGA superclocked in my now about-to-retire from FSX i7 920 system OC'ed @ 4.4Ghz. in a nutshell, I am happy with the 680. It does outperform the 480 slightly in heavy autogen/weather conditions. Even though I use the term "slightly", in the grand scheme of things, there is a very fine line between microstutters and smoothness. So "slightly" translated to what you are seeing on the screen is more like a rather "significant" increase in smoothness.

 

As far as I'm concerned the 680 has the extra bit of smoothness and hosepower under load. As I tend to run unlimited FPS (and that WORKS for me) I experienced just a slight drop against the 480, in FPS maybe by 4-5 frames during scenarios where autogen and weather are not taxing the system. I have some baseline FSX FLT files that I use for consistency in the KSEA area. As far as I'm concerned, if you are in the market for a new card, or building a new system anyway, I see no reason NOT to get it. I can't speak for the 580, and Word Not Allowed has already covered it earlier in the thread. If 680 and 580 perform pretty much identical, then you could save a few bucks and go for the 580.

 

I must have swapped the cards back and forth about four times, just to assure myself that I wasn't experiencing a placebo effect. The latest 301.24 beta drivers are unified, so technically there is no driver swap needed. However, if a given system has never seen a 680, the drivers must be re installed, when the new card is introduced into the system for the first time. After that, you can swap cards back and forth all day long. This makes for a fair apples to apples comparison, as the overall system, and drivers remain untouched. As always, your mileage may vary, as no two systems are ever alike.

 

Now:

I will briefly touch on the experience that I've had with my new Sandy Bridge system. All I can say is wow! This is a clean virgin OS/FSX Acceleration install. OS is on a WD Caviar Black 1.5Tb, FSX is on a WD Velociraptor 600Gb. I have only implemented three tweaks to the CFG. AM-14, HIGHMEMFIX, and UsePools=0. That's it. OS and FSX drives have been tuned/defragged as per NickN's guides as per usual. FSX settings are as per NickN as well but with more aggressive weather settings. For the first time ever in the five FSX rigs that I've built, I can run at a maximum cloud distance without a significant performance hit. On a side note, when you can complete the "Tokyo Executive Transport" mission, without being distracted by stutters and poor performance, then I say mission accomplished. <----------Did you see what I did there?

 

So far I'm impressed, and as I said I have not even overclocked yet. Anyway, I will leave it at that, as I don't want to derail the thread. If anyone wants more details, feel free drop me a PM. Cheers.

[WebMaximus 372]

 

 

Hi guys:)

 

Sorry for taking so long to actually get back to reporting on the 480 vs the 680 in my i7 920 rig. I was busy building a new i7 2700K system, along with win 7 tuning/tweaking/fresh FSX install. I am REALLY Pleased with the performance of the Sandy bridge, and I have not even overclocked yet! However I will leave that aside for now.

 

Well, as I said I was not expecting a quantum leap in performance regarding 480 vs 680 EVGA superclocked in my now about-to-retire from FSX i7 920 system OC'ed @ 4.4Ghz. in a nutshell, I am happy with the 680. It does outperform the 480 slightly in heavy autogen/weather conditions. Even though I use the term "slightly", in the grand scheme of things, there is a very fine line between microstutters and smoothness. So "slightly" translated to what you are seeing on the screen is more like a rather "significant" increase in smoothness.

 

As far as I'm concerned the 680 has the extra bit of smoothness and hosepower under load. As I tend to run unlimited FPS (and that WORKS for me) I experienced just a slight drop against the 480, in FPS maybe by 4-5 frames during scenarios where autogen and weather are not taxing the system. I have some baseline FSX FLT files that I use for consistency in the KSEA area. As far as I'm concerned, if you are in the market for a new card, or building a new system anyway, I see no reason NOT to get it. I can't speak for the 580, and Word Not Allowed has already covered it earlier in the thread. If 680 and 580 perform pretty much identical, then you could save a few bucks and go for the 580.

 

I must have swapped the cards back and forth about four times, just to assure myself that I wasn't experiencing a placebo effect. The latest 301.24 beta drivers are unified, so technically there is no driver swap needed. However, if a given system has never seen a 680, the drivers must be re installed, when the new card is introduced into the system for the first time. After that, you can swap cards back and forth all day long. This makes for a fair apples to apples comparison, as the overall system, and drivers remain untouched. As always, your mileage may vary, as no two systems are ever alike.

 

Now:

I will briefly touch on the experience that I've had with my new Sandy Bridge system. All I can say is wow! This is a clean virgin OS/FSX Acceleration install. OS is on a WD Caviar Black 1.5Tb, FSX is on a WD Velociraptor 600Gb. I have only implemented three tweaks to the CFG. AM-14, HIGHMEMFIX, and UsePools=0. That's it. OS and FSX drives have been tuned/defragged as per NickN's guides as per usual. FSX settings are as per NickN as well but with more aggressive weather settings. For the first time ever in the five FSX rigs that I've built, I can run at a maximum cloud distance without a significant performance hit. On a side note, when you can complete the "Tokyo Executive Transport" mission, without being distracted by stutters and poor performance, then I say mission accomplished. <----------Did you see what I did there?

 

So far I'm impressed, and as I said I have not even overclocked yet. Anyway, I will leave it at that, as I don't want to derail the thread. If anyone wants more details, feel free drop me a PM. Cheers.

 

Thanks for your detailed report, very interesting since I've been thinking to upgrade to a Sandy Bridge based setup as well currently using an iCore i7-920 @ 4 GHz and a GTX 480.

 

Have been thinking about replacing this with a 2700K and a GTX 580...just trying to understand if the performance boost would be big enough compared to my current setup to justify the cost...

 

Might also add that I'm cooling both the CPU and the GPU with an Aquaduct 360 Mark III XT water cooling system and intend to use the same system for the new Sandy Bridge setup mentioned above if I end up upgrading and if so I'm hoping to be able to reach 5 GHz since most seem to reach 4.5 GHz on air.

[Ben Cap 89]

Have been thinking about replacing this with a 2700K and a GTX 580...just trying to understand if the performance boost would be big enough compared to my current setup to justify the cost...

 

You will see a very large jump in FPS, especially is you overclock. 2700K @4.8-5GHZ + GTX580 will be at least 35-40% jump in FPS.

[WebMaximus 372]

 

 

You will see a very large jump in FPS, especially is you overclock. 2700K @4.8-5GHZ + GTX580 will be at least 35-40% jump in FPS.

 

Ok, that sounds great!

 

Wasn't sure if the jump would be that big since I'm already running at 4GHz and knowing the GPU was never that important in FSX compared to the CPU. With a 35-40 FPS jump I'm very tempted...will have to think it over a bit.

 

Thanks for your input!

[dazz 69]

Ok, that sounds great!

 

Wasn't sure if the jump would be that big since I'm already running at 4GHz and knowing the GPU was never that important in FSX compared to the CPU. With a 35-40 FPS jump I'm very tempted...will have to think it over a bit.

 

Thanks for your input!

 

I'm not sure I would get the GTX580. I've done that because my 480 crapped out and it doesn't seem to handle SGSS that much better.

I'm starting to think that the ROP count is what matters for that and there's no difference between the 480 and the 580 there, just the little extra core clocks.

[Ben Cap 89]

480 and the 580 there, just the little extra core clocks.

Cooler temps, less power consumption, and slightly improved OC'ing.

[dazz 69]

Cooler temps, less power consumption, and slightly improved OC'ing.

 

Yeah, absolutely. My reference 580 runs much cooler than my old Windforce 3 480.

But performance wise I'm not sure it's worth the upgrade.

Look at how the 680 doesn't seem to outperform the 580... it's got only 32 ROP's.

 

Wish I had the 480 todo a side-by side comparison in clouds + 4xSGSS

[Ben Cap 89]

4xSGSS

 

Why in our wonderful world would you EVER need 4xSGSS?!?! :shok:

 

When I fly in intense scenery with NGX, I turn off SGSS and fly with 8X MS only.

[dazz 69]

Why in our wonderful world would you EVER need 4xSGSS?!?! :shok:

 

When I fly in intense scenery with NGX, I turn off SGSS and fly with 8X MS only.

 

4xSGSS looks gorgeous to me. In good weather it's perfectly fine and I didn't get a 580 to run the same AA I was running in my 460.

Anyway, what I meant by clouds + 4xSGSS is an scenario to test the capabilities of both cards at massive AA levels, just to make sure I'm testing the GPU's and not something else. I did a flight to Innsbruck yesterday and stumbled across a thick layer of clouds at FL120 and my FPS went down to the teens with 4xSGSS :D

[Ben Cap 89]

went down to the teens with 4xSGSS :D

 

:shok: :shok: :shok:

The thought of dropping into the teens makes me sick to the stomach.....

 

With you ":D" you are implying that you are celebrating over the low FPS?!?!?! :LMAO:

 

580 to run the same AA I was running in my 460.

 

I only turn the SGSS off when flying into FSDT KLAX. - Easily the most demanding airport I have ever flown into.

[dazz 69]

:shok: :shok: :shok:

The thought of dropping into the teens makes me sick to the stomach.....

 

With you ":D" you are implying that you are celebrating over the low FPS?!?!?! :LMAO:

 

It can certainly screw up an already challenging approach to find yourself in the middle of a thick mist or cloud layer with low FPS, yeah.

Bad planning on my part, but hey

 

I only turn the SGSS off when flying into FSDT KLAX.

 

Do you need to? SGSS is not that much of an issue in complex scenery (unless there's tones of AG) as it is in bad weather. I would think at KLAX it's more about the CPU

[Ben Cap 89]

I would think at KLAX it's more about the CPU

 

Umberto of FSDT explained how KLAX is optimized for high end GPUs.

 

At FSDT KLAX, you are GPU limited even with a GTX580. - Hence my purpose for turning off SGSS.

[dazz 69]

Umberto of FSDT explained how KLAX is optimized for high end GPUs.

 

At FSDT KLAX, you are GPU limited even with a GTX580. - Hence my purpose for turning off SGSS.

 

Interesting, thanks for the link.

[Kosta 122]

At FSDT KLAX, you are GPU limited even with a GTX580. - Hence my purpose for turning off SGSS.

 

I must yet again go against the stream apparently and say that with 2x SGSS I'm *not* GPU limited at KLAX, except if I load whole bunch of clouds...