• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

i7 and Microsoft Flight Simulator X

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Papafox

Member
Joined
Jan 24, 2007
I've been comparing two computers with MSFS X. One has an i7 920 @2.8 and the other has a Q9550 @ 2.8Ghz. Both have 4GB ram and one ATI 4850 video card. The i7 computer is significantly smoother.

FPS is tough to benchmark because sometimes it varies widely while the plane is just sitting on the runway and sometimes you get relatively high frames per second (35-45fps) and yet you can still get stutters.

It's all about smoothness in MSFS X. I'll add a few benchmarks if I can find a valid way to compare them, but in terms of smoothness, Nehalem rules. Perhaps it's the memory bandwidth, perhaps it's just a different architecture, but at the same Ghz speeds, Nehalem acts as if it is 33% more powerful than a Penryn Quad.
 
I heard more than once that both Kentsfield and Yorkfield have problems with FSX, Wolfdale/Conroe will run it smoother. I think the problem is the dual die and the "duct-tape" used to connect them, but then again i never had a 775 quad.

I hope to get around to install/repair WinXp on the 920 soon, i might be able to throw in some subjective opinion of [email protected] vs [email protected] (stock cooler suxx).
 
if anything run fraps, log the fps and do the same run flight on both systems using auto pilot.

FSX is just another poorly coded game, my uncle went from a e4400 stock 8800 368mb card to a q6600 and i put in my 8800GT i had bough to try out and the performance gains were minimal at most.

MS has yet to release a good patch to use multiple cores well like they keep sayihg they will since SP1.
 
Here's one benchmark of sorts:
i7 920 @ 2.8 vs. q9550 @ 2.83
* All sliders full max except traffic, which is zeroed out
* 4GB ram & ATI 4850 video in both
* In B-737 cockpit, full wide view, sittling on default Runway at krno with fair wx in mid-day

q9550= average of 27 fps
i7 920= average of 40 fps

Most games don't see much improvement with Nehalem, but it looks like those who predicted that MSFS X would benefit were absolutely right.

Mr. Guvernment, I'll try the fraps test when I get software installed.
 
an interesting benchmark i was wanting to see, thanks.
However, it would be more fruitful if you could do this more methodically, i mean like most reviews are done. Also take the clock a bit higher, atleast 3.2 Ghz for both, reduce resolution and graphics effects and keep others like Physics at max. (i am assuming this is strictly CPU benchmark)
 
50%! Thats amazing, the most optimistic guys at sim-forums were talking 30-35%. Bandwidth could be an issue in addition to raw CPU power.
 
Blue,
This was done at 1280x1024 to keep the emphasis on the CPU

I'll try for more tests today.
 
Did some quick and dirty testing also, really impressive performance, only rarely dipped below 25fps on my favorite intl airport, thats with UT-X at high traffic settings and payware addon aircraft. The e5200 and even the E8400 @4,2GHz would dip below 16 frequently in places like that, so i would say 30-50% improvement. Also around small airports the improvement was big, where i used to lock FPS at 25-30 i can now lock it at 40, if there is room for even more i use that overhead to further increase quality/traffic settings as an extra bonus.

FS2004 only ran this well on new and hot hardware after FSX was already launched, hardware performance is now getting ahead of software bloat for the first time =)
 
Did some quick and dirty testing also, really impressive performance, only rarely dipped below 25fps on my favorite intl airport, thats with UT-X at high traffic settings and payware addon aircraft. The e5200 and even the E8400 @4,2GHz would dip below 16 frequently in places like that, so i would say 30-50% improvement. Also around small airports the improvement was big, where i used to lock FPS at 25-30 i can now lock it at 40, if there is room for even more i use that overhead to further increase quality/traffic settings as an extra bonus.

FS2004 only ran this well on new and hot hardware after FSX was already launched, hardware performance is now getting ahead of software bloat for the first time =)

WOW! Myself being an avid FSX buff, I think you two have completely changed my mind about Nehalem. It looks like there will be a 965 rig for me this XMAS!

Thanks again for this great news guys.
 
Just did a FRAPs test on a local flight over Seattle in F-18 at 1500 ft & 300 knots over the same route. Nehalem still rules but the margin was a bit lower this time:
Q9550 @2.83Ghz= 17fps average in fraps
i7 920 @ 2.8 Ghz= 21 fps average in fraps

That's a 23% increase in frame rate for i7, which isn't as good as the 50% I saw sitting on the runway, but lots of tests will be necessary to bracket the performance difference.

The i7 ran much smoother, though, and produced a considerably-better flight experience. I suspect the Nehalem will gain ground at higher Ghz, due to better memory handling. I hope to have mine up to 3.8Ghz later in week to do more tests.
 
The i7 ran much smoother, though, and produced a considerably-better flight experience.

I too feel thats the biggest improvement. There will be scenarios where i7 give only a few percent extra, and im ok with that, i rather have the increased performance where it counts. For example advanced payware addon commercial airliner about to touchdown during heavy traffic hours, FPS drops to 5 for a second and you can forget about the perfect landing you were hoping for, one small drop in framerates totally destroys a long and otherwise good flight. If things like that can be avoided then i7 is already worth it.
 
Last edited:
Just did a FRAPs test on a local flight over Seattle in F-18 at 1500 ft & 300 knots over the same route. Nehalem still rules but the margin was a bit lower this time:
Q9550 @2.83Ghz= 17fps average in fraps
i7 920 @ 2.8 Ghz= 21 fps average in fraps

That's a 23% increase in frame rate for i7, which isn't as good as the 50% I saw sitting on the runway, but lots of tests will be necessary to bracket the performance difference.

The i7 ran much smoother, though, and produced a considerably-better flight experience. I suspect the Nehalem will gain ground at higher Ghz, due to better memory handling. I hope to have mine up to 3.8Ghz later in week to do more tests.

Is that a like for like comparison, I mean is your RAM speeds and timings the same on both the Q9550 and i7 setup?
 
The Q9550 uses DDR2 while the i7 920 uses DDR3. Particulars are as follows from CPUZ:
Dram frequencies: Q9550: 534, i7: 563
Timings: Q9550: 5,5,5,15, i7: 8,8,8,20

The DDR3 likely helps the performance somewhat, and it'd be interesting to do the Q9550 test in a machine with DDR3. Nonetheless, since the DDR3 is a forced upgrade to use the i7, and since most computer owners are running DDR2 at present, my comparison gives an idea of the performance improvement for someone who shells out the bucks to upgrade his motherboard, processor, and memory. That is the performance increase that I am personally most interested in seeing.
 
This again is making me extremely eager to push the buy it now button on a 965 rig. Guys what are you using for air cooling on the i7? What options do we have currently available? Also, I know Brolloks had tested 2 sticks as performing better than triple channel memory. Are you guys running 2 or 3?

Regards,
dominick
 
Last edited:
I think that the 965 is a bit overpriced. The best bang for the buck would be the 920 atm, but if you have the cash and want the fastest, then the 965 would be it.
 
I think that the 965 is a bit overpriced. The best bang for the buck would be the 920 atm, but if you have the cash and want the fastest, then the 965 would be it.

I will be cooling with LN2,DI, or phase in the future so the unlocked multi of the 965 is a must for me.
 
Last edited:
Back