But didn't you say this? And with Global High settings:Frames, Time (ms), Min, Max, Avg16290, 300000, 33, 82, 54.300Anyway so I re-did the test with the tweak and it went so much betterMin: 36 Avg: 57 Max:78I have some DDR2 800Mhz ram laying around, will it work with my Mobo?

see post #85, I was doing the test with the wrong settings so my results were far better than they should. The right ones for global high are in the mid 30's (avg)No, your board won't support DDR2, but you can dumb down your system to dual channel by removing 2 of your ram sticks. They have to be both in the same channel and you can check that with CPU-Z, but if you're not sure how to do it, don't worry. You've been a great help already. Very much appreciated.I would have never thought there would be such a big difference between I5 & I7. I'm puzzled now

But didn't you say this? And with Global High settings:Frames, Time (ms), Min, Max, Avg16290, 300000, 33, 82, 54.300Anyway so I re-did the test with the tweak and it went so much betterMin: 36 Avg: 57 Max:78I have some DDR2 800Mhz ram laying around, will it work with my Mobo?

Clarkdale vs Lynnfield vs BloomfieldThanks guys for your testing. So in summary, KSEA in the CRJ on autopilot in global high settings using the link above we have:(Key: CPU = min/avg/max FPS, graphics card)i3-540 @ 4.4 ghz = 20 / 36 / 51 (8800GTS 320mb)i5-750 @ 4.0 ghz = 21 / 36 / 57 (GTX460 1gb)i7-930 @ 2.8 ghz = 27 / 45 / 73 (GTX470)CPUS compared: http://ark.intel.com...473,42915,41447, I would also say that the systems are appropriately matched with a video card although I'd be interested to see the i7 with a GTX460. We've tried to standardise our testing as much as possible, but at least we have an idea of what the different systems can do.Will the i7-930 eventually be overclocked?

can you guys please post your bandwidth & latency results in maxxmem?http://www.brothersoft.com/d.php?soft_id=270353&url=http%3A%2F%2Ffiles.brothersoft.com%2Futilities%2Fmiscellaneous%2FMaxxMEM2_preview.zip

Clarkdale vs Lynnfield vs BloomfieldThanks guys for your testing. So in summary, KSEA in the CRJ on autopilot in global high settings using the link above we have:(Key: CPU = min/avg/max FPS, graphics card)i3-540 @ 4.4 ghz = 20 / 36 / 51 (8800GTS 320mb)i5-750 @ 4.0 ghz = 21 / 36 / 57 (GTX460 1gb)i7-930 @ 2.8 ghz = 27 / 45 / 73 (GTX470)CPUS compared: http://ark.intel.com...473,42915,41447, I would also say that the systems are appropriately matched with a video card although I'd be interested to see the i7 with a GTX460. We've tried to standardise our testing as much as possible, but at least we have an idea of what the different systems can do.Will the i7-930 eventually be overclocked?

Exactly. don't forget that my CPU is only at 2.8Ghz while you guys are in the 4Ghz. That at least makes up for the GPU, Im pretty sure.@Dazz yea I'll test my memory, I'll have the results this eve.

Thanks a lot Luckyman.I tested my memory and FSX:1600GHz 9 - 9 - 9 - 24: BW = 14,5GB/sec..................51ns (latency)36 FPS_______________________________________________1200GHz 11 - 11 - 11 - 31: BW = 13,5GB/sec..................63ns (latency)34 FPSSo for a 7% memory band width increase, there was a 6% frame increase. That could very well mean that tripple channel makes a good difference (ram speed itself doesn't really matter much except for the fact that it might limit overclock if you can't set a low RAM multi)Actually Luckyman, memory results also depend on clock speed, so we'll have to keep in mind you're on stock clocks as you said

At stock:Overclocked:PS just saw dazz's post. Will go back and do the memory vs FPS thing. I am also looking at the i7's double bandwidth QPI link with jealousy.

That shows the weaker IMC of the I3. @ 4.4GHz, almost the same speed and much better latencies the BW is aprox 25% lower than my I5.What I don't understand is that if memory is the limiting factor why we get such similar results in FSX. Might be addons. I'll restore my default textures, disable UTX & GEX and repeat the tests

The constants: i3 @ 4416mhz, 8800GTS 320mb video card, Vanilla FSX SP2 (no add-ons), OCZ Solid 2 60gb SSD, KSEA flight in CRJ at "global high.cfg". Memory timings 7-7-7-21.All I did to change the memory multiplier from 2:8 to 2:6 and this is what I got:Memory @ 1536mhz (8x192mhz) : 22 / 41 / 53(min / avg / max FPS) (redone)Memory @ 1152mhz (6x192mhz) : 23 / 39 / 51 (redone)RESULTS UPDATED 40 MINUTES AFTER ORIGINAL POST. Interesting. (Initially I thought it was much worse with slower memory speed but I got the settings wrong again)

Edited September 8, 201015 yr by bangoman

It's actually just a 7% bandwidth difference. Wow, I can't believe it makes such a big performance difference.I think I'll test single vs dual channel instead then

Hang on I messed up the settings in my original run and subsequent posting with that HUGE difference. Error fail. The results are basically the same. Sorry for misleading, post edited.It's so hard to get consistency with FSX, I can see why it's not used as a benchmarking tool.

Edited September 8, 201015 yr by bangoman

Hang on I'm editing my findings as I messed up the FSX settings. It's running on my desktop right now.

Ok, I missed that. BTW, are you using the BP=0 tweak please?

LOL, how did you get the forum to display edit logs?

Not using anything like bufferpools tweaks - the guides I linked to above said not to bother and I've got no complaints about my setup. I'll google and look into it. Are you using it?

yes I am. It's just to make an apples to apples comparison