http://forums.simflight.com/viewtopic.php?t=56910Marco

Are you kidding me? A C2D X6800 would eat a P4 3.2 for breakfast in FSX FPS, not be level with it. Somebody needs to run an actual benchmark with the game rather than speculate based on synthetic benchmarks.Gary

>Are you kidding me? A C2D X6800 would eat a P4 3.2 for>breakfast in FSX FPS, not be level with it. Somebody needs to>run an actual benchmark with the game rather than speculate>based on synthetic benchmarks.>>GaryFunny my experience is total different maybe he used the default setting which is due to FSX detecting the C2Duo wrong...With my E6600 I blow all that figures from the table lolMost unrealistic benchmark I've ever read ;-)Just garbage...Andr

Wow. This is great. I always knew my old OC'd P4 3.2 was way better than an EX6800 and now I have the proof. What a great set of benchmarks! No need to upgrade my CPU now. Maybe I'll spend the money I just saved to upgrade my 32MB video card. Anyone seen any benchmarks like these for video cards? Maybe I can save even more money.DougEdited 'cause I can't spel

I don't know how this benchmark was made but I must agree from my own observations that core clocks are very important for FS-X performance.Genrerally the higher you core is clocked the beter performance you get. Where this goes wrong is when you compare P4's with Core2 Duo's only on clockspeed. The Core 2 Duo is more efficient then the old P4 core so you get more performance with the same clockspeed there.

>Are you kidding me? A C2D X6800 would eat a P4 3.2 for>breakfast in FSX FPS, not be level with it. Somebody needs to>run an actual benchmark with the game rather than speculate>based on synthetic benchmarks.>>Gary Gary, don't forget something -- FSX is NOT multi threaded very well at all. I have an X2 and a G15 keyboard that shows a cpu monitor for both cpus during the game. One cpu is maxed out, the other is doing nothing substantial (like say 5% usage, possibly windows tasks, maybe some data shuffling around, who knows)...If I get the marketing right... an X6800 means its a dual "3400" which technically means amd says it is on par with an intel running at 3.4. Given that FSX uses barely a single cpu, I would say that a P4 3.2 and the X6800 won't be very very far off. And I'm going by AMD's own hyped up marketing numbers!Until FSX gets a patch to try and use the other cpu cores for SOMETHING, dual cores and smp machines just won't get the benefit they deserve.Oh, and to those who say that it is just too difficult to write games to use multi-cores, if you play most recent games, they really are using both cores. Maybe not in the most efficient way, but they are.

>If I get the marketing right... an X6800 means its a dual>"3400" which technically means amd says it is on par with an>intel running at 3.4. Given that FSX uses barely a single cpu,>I would say that a P4 3.2 and the X6800 won't be very very far>off. And I'm going by AMD's own hyped up marketing numbers!Not quite. First off, the x6800 is an Intel part and they don't do any sort of "performance rating" based part numbering scheme like AMD. It's simply just a model number.Second, the AMD Performance numbering model scheme is based on a comparison against the original Athlon (1st gen K7 core), not on the Pentium4. And it's calculated based on a suite of benchmarks that AMD manages to test a wide range of calculations and features in the chips. I would NOT consider those numbers hyped up. They are very much inline with what you can expect in most cases (some benchmarks will run better, some will run worse, but on average you get the performance it says you should relative to an original K7).Now, to your last point, a X6800 will blow the doors off a P4 3.2! The P4 design is a pig - always has been. Don't think for a second that a x6800 won't kill a P4. The problem here is that the game only looks at the CLOCK SPEED to determine performance (which is a no-no!). At the same clock speed, the Core2 Design can perform somewhere around 50% more calculations/work than the P4 design. So even if you'relosing around 600Mhz on the x6800, you're MORE than making up for it in core efficiency. >Until FSX gets a patch to try and use the other cpu cores for>SOMETHING, dual cores and smp machines just won't get the>benefit they deserve.Well, yes and no. Having multiple cores DOES allow you to offload a lot of the background OS processing to the 2nd core, freeing up additional cycles on the 1st core for the game to use. As a result you get about a 5-10% increase in performance going dual core over an identical single core CPU.>Oh, and to those who say that it is just too difficult to>write games to use multi-cores, if you play most recent>games, they really are using both cores. Maybe not in the most>efficient way, but they are.True. And I'm sorry, but I write multithreaded code all the time. I think it's BS that they say they can't do it with games. It can be done, it just requires thinking about the main game processing loop a little differently. The real issue here is that the % of multi-core CPU's out in the wild vs. single core CPU's is low right now. If they write the games to really take advantage of multi-core, then the games won't work too well on single core and that alienates a large percentage of potential customers. This is a case of the software guys having to wait on the hardware to reach critical mass before switching over.--2002cbr600f4i

>>Are you kidding me? A C2D X6800 would eat a P4 3.2 for>>breakfast in FSX FPS, not be level with it. Somebody needs>to>>run an actual benchmark with the game rather than speculate>>based on synthetic benchmarks.>>>>Gary>> Gary, don't forget something -- FSX is NOT multi threaded>very well at all. I have an X2 and a G15 keyboard that shows a>cpu monitor for both cpus during the game. One cpu is maxed>out, the other is doing nothing substantial (like say 5%>usage, possibly windows tasks, maybe some data shuffling>around, who knows)...>>If I get the marketing right... an X6800 means its a dual>"3400" which technically means amd says it is on par with an>intel running at 3.4. Given that FSX uses barely a single cpu,>I would say that a P4 3.2 and the X6800 won't be very very far>off. And I'm going by AMD's own hyped up marketing numbers!>>Until FSX gets a patch to try and use the other cpu cores for>SOMETHING, dual cores and smp machines just won't get the>benefit they deserve.>>Oh, and to those who say that it is just too difficult to>write games to use multi-cores, if you play most recent>games, they really are using both cores. Maybe not in the most>efficient way, but they are.>This takes the award for 'Best Lack of Understanding Regarding CPU Design' !!Core 2 Duo, regardless of software being multithreaded, is the fastest CPU available today. Single threaded apps, multi threaded apps, it doesn't matter.No previous P4 or Pentium D can even hope to compete with an X6800 or an E6600 for that matter. The E6600 already beats an FX-62 and that was the fastest available at the time of the Core 2 Duo launch.Folks, the CPU Mhz race is over. Core 2 Duo is many times more efficient at data throughput than any previous Intel design, hence it doesn't NEED higher MHz to do so. We should all be thankful that Intel CPU's are not mini space heaters anymore!2 cores running at 2.4 Ghz each, but even ONE of those cores running a single thread app will still beat any previous gen CPU, including the FX-62 and any Intel P4 Extreme Edition.

Exactly,I will take my E6400 @ 2.66 Ghz against ANY P4 3.8 ghz or FX-62 anyday of the week.

This is probably one of the better ways to compare the single threading performance of the cpu's:http://www23.tomshardware.com/cpu.html?mod...2=464&chart=188P4's aren't even close and the FX-62 performs better than you would think.Bruce