So... how does that 737NGv2 look now? :)

May have spoken too soon here...After spending some more time in FSX, I started to get the feeling that somehow it wasn't even as fast as the GTX260 had been in the sim. Turns out I was right, I've swapped between both cards multiple times tonight and the 260 is very clearly faster somehow with the exact same settings in FSX and in the driver/Inspector. I'm kind of at a loss here because other games work fine on it and are definitely performing much better than the 260 was. My 3DMark 11 score and Metro 2033 benchmark score were both right in line with other early 570 adopters at the hardware site forums, which suggests that the card is fine and the problem is somehow FSX specifically in combination with this card. Power draw never got above 500W even running those really advanced new games or benchmarks. Sigh... this one might be going back guys.

May have spoken too soon here...After spending some more time in FSX, I started to get the feeling that somehow it wasn't even as fast as the GTX260 had been in the sim. Turns out I was right, I've swapped between both cards multiple times tonight and the 260 is very clearly faster somehow with the exact same settings in FSX and in the driver/Inspector. I'm kind of at a loss here because other games work fine on it and are definitely performing much better than the 260 was. My 3DMark 11 score and Metro 2033 benchmark score were both right in line with other early 570 adopters at the hardware site forums, which suggests that the card is fine and the problem is somehow FSX specifically in combination with this card. Power draw never got above 500W even running those really advanced new games or benchmarks. Sigh... this one might be going back guys.

Yeah, there is something wrong, either with it, or with your OS or settings on this one. Very strange. Did you try setting it with Nvidia inspector? Did you try cranking the settings way up with the Nvidia SLI tool using the high profiles like 16 x AA? Both work well with the 580 GTX. Let us know what you decide to do.Kind regards,

I as well have a 620HX and am very interested in the GTX 570.I'm interested in hearing if it is confirmed that the PSU is enough for the card.I have a Kill-A-Watt meter myself, I may do some testing to see how much my system is currently pulling.

Yeah, there is something wrong, either with it, or with your OS or settings on this one. Very strange. Did you try setting it with Nvidia inspector? Did you try cranking the settings way up with the Nvidia SLI tool using the high profiles like 16 x AA? Both work well with the 580 GTX. Let us know what you decide to do.Kind regards,

Yeah, it's unbelievably strange to me because the other games are working fine - it's seriously just FSX that's slow. I read your review of the 580 vs. the 275 and 480 and this doesn't look a thing like the numbers you're seeing. Using Inspector and everything, of course yeah.What P score do you get in the new 3DMark11? I get a P5235 with it, which seems in line with what I saw other people with 570s getting on the hardware sites. (some were a bit higher, like in the 5600 or so range, but they also had better CPUs)

I hope problem solve soon , I am a little bit confused! I am going to upgrade from 285 to 570 or 580. So I hope Ryan can reach to SpiritFlyer results.

What P score do you get in the new 3DMark11? I get a P5235 with it, which seems in line with what I saw other people with 570s getting on the hardware sites. (some were a bit higher, like in the 5600 or so range, but they also had better CPUs)

TabsI got P6777 whatever that means. I don't know if that is good or bad because I never used it before. Where can I see other peoples scores?Kind regards,

That's a very good score for a single 580 Stephen, guys with it at stock are seeing around 6100-6200. The actual 3DMark search site is apparently broken, I'm getting numbers off the thread at the HardOCP forum:http://hardforum.com...d.php?t=1567233What sort of settings tweaks did you have going in FSX for the numbers you got in your big review post? I've swapped the cards two more times now today and I'm absolutely positive FSX is far slower with the 570. The 570 is however getting almost 3X the performance in any other game or benchmark, it's so bizarre. I ran the Unigine Heaven DX10 Benchmark http://unigine.com/products/heaven on both cards - the 260 got 32FPS, and the GTX570 87FPS. This makes absolutely no sense to me that there's such a massive positive difference in other games, yet FSX is slower.

I ran it again but this time with hyperthreading turned on and this is what it came up with. Nothing extraordinary I don't think. Tabs there is nothing out of the ordinary in my CFG file other than a few tweaks and settings I have picked up over the years that work as opposed to the many that don't. There is the Bufferpools reject threshold, highmemfix, 6.5 LOD radius and other odds and ends. The trees and buildings tweak was removed for the tests so everything was Kosher, with nothing that reduces detail or quality. The Nvidia settings were set as shown in Nvidia Inspector. There is nothing in my rig that is not common. Most people I know in the Prop Club and in the FTX forums that have similiar machines have similiar performance, some a bit better, some a bit less, but nothing like what you are describing. There is something mismatched or bottlenecking your performance in FSX with that video card, but I have absolutely no idea what it is or where to start looking. The answer however does not lie with the 260, that's for sure. Good luck.Kind regards,

I just went out to Fry's and replaced the PSU with a 775W 60A 12v rail model - absolutely no change. It's not the PSU.I really think this is probably somehow a driver issue with FSX specific to the 570 now, it's the only thing that makes sense. Other games work fine so there's probably nothing wrong with the card - I don't know how a problem with the card would manifest in only one older application and not in all these new games that push it so much harder.The only other thing I can think of to try is reformatting and clean installing Win7, but I really don't feel like doing that. Gonna return the card on Monday and just stay on the 260, I can't afford to have 15-20 FPS instead of locked at 40 for the NGX beta test, which is what I'm seeing between the two cards right now...

Update:I think I'm starting to understand what's happening now after running some more tests today. The Fermi cards appear to scale differently with increased CPU clockspeed than the older Nvidia cards do. I pushed my CPU overclock a bit further to 4.0 this morning and I saw a pretty substantial framerate increase in FSX as a result. The performance with the 260 in went up very slightly, but nowhere near the increase I saw with the 570 in. I think the card somehow gets fed a lot more data to work with or something to that effect when the CPU clock is higher. It's doesn't appear to be a linear relationship either - the performance goes up by a ton with small increases in clock at the top end of the CPU's range. At 4GHz it's much better than it was at 3.8, enough that I'm considering keeping the card now and just getting a Megahalems or something like that to try and get it to 4.2+.Stephen - I'd be curious to see what your results with the 580 are if you downclock your CPU to 3.8 like mine was...

Hi Tabs,The more you remove impediments, or unclog the system the better. No piece works, particularily the GPU in isolation, as it needs to be fed. This posting I made awhile back on the ORBX FTX forums might explain it better. If you follow down the thread you will see updated information as well that should confirm the importance of overclocking your CPU as long as it does not race past the other conponents in your system. The Value of Overclocking an Integrated SystemI don't raise or lower my overclock for tests. Once a system has reached a very high overclock and is operating 100% stable it is best to leave it there and not jack the settings around. The clock speed is not the only factors that are adjusted so I only have two profile settings, 4.3 and the default 2.8 profiles. In case you are interested here is a guide for overclocking that goes into some specific details showing my system and a few others:Bios Overclocking ExamplesThe Megahalems is still the best according to this benchmark review. It sure is a good looker!Best Air Cooling Solution BenchmarksKind regards,

Thanks Stephen,Which fans did you get for your Megahalems? I think I'm gonna order one tomorrow.

Sounds to me as if you got a bottleneck in the system... there is no way you should get the same results unless you have a bad configuration. My guess you have settings that are now just being applied because your hardware actually supports it....

Sounds to me as if you got a bottleneck in the system... there is no way you should get the same results unless you have a bad configuration. My guess you have settings that are now just being applied because your hardware actually supports it....

You know Veeray, that is an excellent point. I have never thought of that before, but that would be a perfect mask ohiding increased performance in a lot of cases!Kind regards,