JohanI thought the same until I bought my new system using a e6600 dual core CPU @ 2.4 more than DOUBLED my frame rates in FSX compared to my old p4 northwood @3.3 gig, these new chips can execute more instructions per clock cycle.There's nothing to touch these new Dual core( conroe) chips ATM.John

Valk: like i said mate, keep telling yourself that...Until there is a patch then there's no way that this game will run at 30fps in a major city with big iron or addons. I am building an E6700 machine instead of a quad core because I'm disgusted at the price gouging going on with this cpu so I will wait for 6 months until they are more reasonable and just o/c the e6700 to 3.6ghz. The quad cores actually are a tad slower than the 6700 and 6800 cpus anyway but i wanted it for multimedia apps and for Alan Wake which will benefit greatly from the extra 2 cores. I am under no illusions that this new system will run FSX the way that I would like but it should provide for at least a 25-33% increase in frame rate. When you're getting 15fps in a major city then you go to 19 or 20 which is still really not that optimal. My hopes are pinned to DX10 and the patch...

>>People are already reporting 20 FPS with>>sliders right on current machines. >>Where? Can you please post a link.>>GaryHey Gary, it is around here somewhere. I have seen at least two posts in the last week or so talking specifically about this on THIS forum. If I find it, I will let you know, but these rigs were EXPENSIVE, and were not home built. I think I can cut down on a lot of the costs of building a new computer by essentially stripping the unneeded things. For example, if I switch to a new SATA drive, there is no reason to have 400GB of harddrive space for me. I could easily save myself a few hundred dollars on that alone since I have NEVER taken up more than 70GB on any HD.Just my $.02. If I find one of those posts before you, I will update it on this post.

hmm, so in a few months time the top dual core now will cost less, and might be a good choice. an extra core wouldnt harm.So, the mhz's dont say anything at this moment, a 2.4 does better than a 3.3..Puzzled, but learning.JohanA LITTLE LESS CONVERSATION, AND A LITTLE MORE ACTION PLEASE

Hehe, there are so many confusing factors when considering a PC CPU that will work for today. I never plan for tomorrow, because tomorrow rarely works the way I intended... I plan for today. Buy something for tomorrow and you may find that technology has changed. A nice example of this is the 64 bit first generation processors, which gave little to no help for those that invested in them.While a lower MHz "can" produce better performance, it all greatly depends on the actual architecture of the chip. There are so many minute factors that make a large difference when it comes to selecting the correct chip for you. Like I said, head over to PCStats and review different chip designs and find out what works for you and your budget. You specifically want to pay attention to the actual benchmarks for each processor, as compared to other quality top of the line processors. You also need to compare those to older chip standards like AMD3000+ series.

Johan,I suggest reading this article. I think it will bring you upto speed.Glennhttp://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795

>Johan,>>I suggest reading this article. I think it will bring you upto>speed.>>Glenn>>http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2795Not a bad article. Thanks for the link.

I agree, I just got the e6600 with a 7900gs and frame rates have doubled in every situation if not tripled. This is coming off an amd athlon xp 3000 and a 6600Gt. I will note that I'm running out of memory as FS is using every last ounce of my 1gig even with using FSautostart. I have another 2 gigs on order.I was happy with the low settings on my old computer, but now I've opened pandoras box and can't help but try to run it as high as it can go. Makes me understand why people are complaining, turn the settings down? Never!I think the reason the core 2 duos are faster despite lower clock speeds is, as has been stated, the design of the chip. They are much more efficient at processing the data at lower speeds than the high cycle single chips.I'm not an expert, but Tomshardware.com's benchmarks and the benchmarks in the posted article all favour the core2duo's by a large margin for the most part. Anyway, I noticed the same thing with FSX and the 2 cpu's running per the original post, is that just a phantom image? I didn't notice any increase in performance but I've been fiddling so much with settings that I haven't done a test.Ian.

I will be interested to see what links you come up with, as the ones I have read show that FSX can be easily brought down to single digit FPS, even on the most beastly setups out there (3.9HGz C2D + 8800 GTS graphics and ample high speed RAM springs to mind). Heck, even my own experience, albeit with beta 3 not the release version, can produce 6-7 FPS at KSEA with not a single addon loaded!Gary

Sorry,But I would wager the performance of a well built home rig to be superior to that of anything which Dell or Alienware can offer.They only have access to the same parts you and I do........(in fact, they usually use lesser quality items such as generic motherboards built for Dell with no overclocking features etc)

>The quad core processors won't improve your frame rates>period from what I understand. The actual clock cores are>lower than top in dual core processors, which leads me to>believe you could actually suffer from worse performance!>Well, not necessarily.The Kentsfield may well be a faster architecture than the Core 2 Duo.If that is the case, it will be faster in FSX. But if it's a slower architecture, with some extra cores, then yes, you would be correct, it would be slower in FSX.As you know already, you can't rely on the clock speed of a processor to determine how fast it might be.As an example, a 2.13 ghz E6400 Core 2 Duo is much faster than a 3.8 ghz Pentium 4. If you clocked a P4 at 2.13 ghz, it would pretty doggy .The problem is, FSX won't take advantage of the other cores in the Kentsfield very much. But a single Kentsfield core will *possibly* be faster than a Core 2 at the same clock speed. I have not checked any test places yet to see though. I would hope it would be, else it's no improvement at all for us FS'ers.RhettAMD 3700+ (@2530 mhz), eVGA 7800GT 256 (Guru3D 93.71), ASUS A8N-E, PC Power 510 SLI, 2 GB Corsair XMS 3-3-3-8, WD 250 gig 7200 rpm SATA2, CoolerMaster Praetorian

I'd like to politely disagree with this statement. As usual, perf isnt that simple. And there are several issues lumped together here.1)There is the FSX CPU side performance, and how multicore and threads and application architecture affect that.2)There is FSX GPU performance, and how FSX D3D API usage affects that.3)There is API performance and how both DX10 and the improved driver model on Vista affect application performance.Performance in 2 and 3 is somewhat orthogonal to 1, so its just not correct to state that 2 and 3 are completely dependant on the applications threading model.In terms of 2 and 3, The driver model on Vista has moved a lot of functionality out of kernel mode, that means the driver can avoid costly Ring3-Ring0 transitions in a lot of cases now that it couldnt on XP. Even for DX9. Then there is D3D10 and its major improvement of Draw call overhead and handling the "small batch" problem. Given early testing on real hardware it looks like the new driver model and D3D10 are delivering on these promises.In terms of 1, one of our biggest issues is the sheer number of Draw calls per frame that we issue. And we stall in the driver on the CPU due to them. Simply moving those Draw calls to another thread will not do anything to change the cost equation. We need to look at better batching or instancing to reduce the number of calls there. Looking at the current state of FSX re perf, understanding where the bottlenecks are, and investigating possible fixes is all part of what we are doing right now. As I said in the infamous "Lost Thread", I will be back in a couple weeks to talk more about what we have determined thru our investigations.But please, dont imply that the only fix is to do something about threading, its just not that simple.

Phil,I appreciate your explanation.So maybe Vista and DX10 will after all, produce some major gains once you have had chance to digest it and develop the DX10 patch?Looking forward to hearing more in a couple of weeks.....

I can definitely state that we are looking at a mix of visual features and using the API more efficiently for the DX10 patch; so it wont be just new features that cost more frame rate.

^Wooohoo!