Hardware Comparison Test

[Dougal 165]

I didn't really know where to post this, but suppose here is most appropriate....  I thought some of you guys might find it interesting.

 

I recently sourced the parts, and was about ready to pull the trigger on a new Skylake build. Until that is, I tried this experiment…

 

A nearby neighbour has the following PC:

 

Chipset:               Asus Z170 Deluxe

CPU:                      i7 6700 @ 4.6ghz (OC)

RAM:                     16Gb Corsair Dominator (3200mhz)

GPU:                     nVidia GTX1080

 

My own current rig, is just beginning to creak and groan a little, even though it’s only about 5 years old:

 

Chipset:               Asus Z68 V-Pro

CPU:                      i7 2700k @ 4.6ghz (OC)

RAM:                     8Gb Mushkin Black (1600mhz) DDR3

GPU:                     nVidia GTX970

 

 

This was a deliberate ‘apples to oranges’ test, to compare two different systems.

 

Both PCs are running fresh installations of Windows 7 (x64), albeit one ‘Pro’ and one ‘Ultimate’, clean installations of FSX-SE and P3Dv3.4 and very little else.

 

Testing was only done at 1920x1080 as we both only have 24” monitors.

 

I’m not technical enough (or well enough) to go into exact numbers, or to know exactly what to test, in order to make this in any way, a proper scientific test. To be honest though, with FSX & P3D, I believe it’s often in the eye of the beholder, rather than an exact science.  We did manage to get both machines running the same flight, side by side, which was sort of fun;-)

 

I was REALLY surprised by the visual results!  When running either a default Mooney, or the NGX; a hi res addon scenery, or a default airport, the difference between the two systems when running FSX or P3D ‘look’ negligible!  I was often hard pressed to see ANY difference.

It became a little easier to spot the differences when AA & AF were cranked up, when the Skylake PC definitely seemed crisper and with less shimmering.  I’m guessing that had much more to do with the 970 graphics card reaching its limits, rather than anything CPU or memory wise.

 

Not worth going into all the settings, as both were always identical.

 

Both PCs are ALMOST silky smooth with default aircraft. The Skylake PC has perhaps VERY slightly less in the way of micro stutters when first panning in outside view. Could this actually be due to a bigger and faster SSD, rather than actual better CPU performance?

 

Locked at 20fps with the PMDG NGX, both machines were still performing well, and with very few stutters.

Locked at 30fps, my rig hit the wall first, but only JUST.  It often doesn’t make the 30fps, topping out around 24-28 and has slightly more exaggerated stutters when changing view.

 

Set to unlimited FPS, the Skylake PC was certainly peaking slightly higher, and so resulted in a higher ‘average’ frame count.  A bit like benchmarking though, I feel this a little deceiving, as the actual visual appearance and overall behaviour is almost identical.  The extra few frames when unlimited, don't actually improve anything.

 

I could probably say a whole lot more, but this sort says enough.

 

I think for now, I'll hang on to my setup AND my cash;-)

[ArjenVdv 190]

24-28 FPS seems low on both systems. My system is similar to your current system, even though my GPU is worse. In default scenery, and in the air, FPS is 40-60 in the NGX. Taxiing at add-ons airports, FPS is at least 25. But I'm running BP=0 so maybe that's where the difference is. 

 

Nevertheless pretty disappointing though. I have seen people reporting different results, saying their new system runs FSX so much better.

[ryanbatc 12,803]

Locked fps are always much lower than unlimited.  If he's locked at 30 and get's 24-28 that's not surprising.  Unlocked I'd guess it would be 40+

[Dougal 165]

The point I was getting at, is the very little 'visual' difference between the two.

[martin-w 3,234]

deleted

[martin-w 3,234]

 

 

Set to unlimited FPS, the Skylake PC was certainly peaking slightly higher

 

 

 

How high? What percentage increase over the 2500K system?

 

Skylake is undoubtedly faster in single core performance than the 2500K, and even faster still multi-core. 30% greater single core performance. That doesn't mean you'll get 30%, but clearly the increase in frame rate should be significant.

 

 

http://cpu.userbenchmark.com/Compare/Intel-Core-i7-6700K-vs-Intel-Core-i5-2500K/3502vs619

 

"Visually" not better you say. But the point is that newer, superior CPU architectures provide improved IPC, and sometimes higher frequency too. 

 

For your CPU upgrade think frame rate rather than "Visually" better. Of course, if that higher frame rate enables you to increase slider settings then yes, you will experience a better "visual" experience.