My i7-2700k to i7-8700k upgrade experience!

[ahsmatt7 1,199]

Hello!

I have never done one of these benchmark style write ups so I apologize if my method is extremely crude and not at all scientific. I just wanted to give an idea of performance.

 

 

I have been using a computer that has had 8GB of RAM, a GTX 1080 ti, a 2700k OCed to 4.8GHZ on air and conventional hard drives for the past 6 years.

I finally decided to build a whole new computer to take advantage of the hype surrounding the 8700k. Here is my experience!

 

Rig:

NZXT S340 ELITE case

Gigabyte Auros Z370 Gaming 7 motherbaord

i7-8700k Oced to 5.0 GHz @ 1.305v

Kraken x62 AIO

GTX 1080 ti @4k

Corsair Vengance RGB 16GB @3000MHz CL15

Samsung 960 EVO NVMe M.2 SSD 500GB (P3d v4.1 and W10 64-bit boot drive)

Samsung 850 EVO SATAIII 250GB storage drive

Corsair CX 850w PSU 

 

I have always found that the KSFO area has always been the place that taxes my sim the most. I know most agree that KSEA is the place but for me, its always been KSFO. I decided to do my testing with the weather theme "Heavy Thunderstorms" in order to put clouds and precip in the mix.  I also went with night time as the time and day in order to test Dynamic Lighting performance. I used PMDG 747 QOTSII and two airports. FlightBeam's KSFOHD v2 (airport Dynamic Lighting OFF) and FlightBeam's KDENv2 (airport Dynamic Lighting ON). I used different Dynamic Lighting settings for these 2 airports to show the performance hit when having that option selected. 

I decided to go with these P3d.cfg tweaks.

[TERRAIN]

TEXTURE_SIZE_EXP=9

TEXTURE_SIZE_EXP=10 was awesome to see but I noticed that autogen was pooping up in large square chunks about 10-12 miles in front of me and looked really bad. IT also slowed down loading times significantly. During my use of level 10 of this tweak, I noticed VRAM usage hovering around 8 GB. 

 

I do have the NORCAL ORBX region installed at the moment as well. I deselected all the options that have a "Larger impact" on performance in the ORBX configurator. These included wind farms, Oakland bay area ships and cranes as well as docks and marinas.

In most situations, the hardest time our sim has is when there is bad weather and while we are at a detailed add-on airport; as well when we are on final to said airports. With the introduction of Dynamic Lighting, that has thrown in a variable that cripples performance in most situations. 

All this being said, I did 8 scenarios...(weather theme of "Heavy Thunderstorms" and Night time)

on runway 28L @ KSFO with all light OFF and then with all lights ON

on final to runway 28L @ KSFO with all lights OFF and then with all lights ON

On runway 26 @ KDEN with all lights OFF and then with all lights ON

On final to runway 35L @ KDEN with all lights off and then with all lights ON

 

Here are my settings I used in P3D v4.1

 

Here are pictures of the First two scenarios. (the FPS counter is on the top left)

Lights OFF...

Lights ON...

Lights ON...looking away from the main terminal buildings and having Lights out of view...

Spot Plane with Lights ON...

Here are pictures of the second two scenarios. (FPS counter on top Left)

Lights OFF...

Lights ON...

Here are pictures of Third two scenarios. (FPS Top Left)

Lights OFF...

Lights ON...

Lights ON...looking away from main terminal buildings and having Lights out of view...

Spot plane with Lights OFF...

Here are pictures of the Fourth two scenarios. (FPS top left)

Lights OFF...

Lights ON...

 

 

I topped out at around 8.2 GB of RAM and I saw VRAM usage hover around 3.8 GB. (VRAM usage was a lot lower than I expected it to be.) 

CPU temps hovered around high 60s and at times, crept into low 70s. CPU usage stayed around 70-90%

 

 

I decided to show the difference in performance between two different airports with two different Dynamic Lighting settings in order to show how the lights hurt performance. 

What surprised me was that at KDEN, I barely noticed a performance hit with both the airport and the airplane having Dynamic Lights on. However, at KSFO, I noticed a big performance hit with ONLY AIRCRAFT Dynamic Lighting on. 

I also noticed that having KSFO airport Dynamic lighting on brought my FPS down to 15fps with all aircraft lights ON and it hovered around 20 FPS with all aircraft lighting OFF.

As a side note, I did not notice a performance hit at all with all aircraft lights ON at LatinVFR's KSNA while using all PMDG airplanes.

It seems as if the performance hit from the Dynamic lights depends on how the airport was developed???? Who knows. 

That's all I have for now. I'm sorry that this was such a rough test but I just wanted to show what the 8700k does fro P3d v4.1 as well as show those of you who have a 2xxxk i7 what the upgrade will bring if you do decide to upgrade to an 8700k.

 

 

[Tokitaumelie 164]

Thank you for this.

Can you confirm resolution for all above was at 4K? Do you think there would be any improvement if you were to lock frame rates at around 20-30FPS with Vsync and TB enabled?

Were the differences in DL due to ORBX NORCAL?

Planning for an upgrade next year and this helps very much.

Kind regards

Edited November 8, 2017 by Tokitaumelie
NorCAL Query

[ahsmatt7 1,199]

1 minute ago, Tokitaumelie said:

Thank you for this.

Can you confirm resolution for all above was at 4K? Do you think there would be any improvement if you were to lock frame rates at around 20-30FPS with Vsync and TB enabled?

Planning for an upgrade next year and this helps very much.

Kind regards

Yes these were all at 4k.

I played around with locking fps and it held 30fps all the way around but with the dynamic lighting on at ksfo, it didn't make a difference. I still got the same performance.

[Tokitaumelie 164]

1 minute ago, ahsmatt7 said:

Yes these were all at 4k.

I played around with locking fps and it held 30fps all the way around but with the dynamic lighting on at ksfo, it didn't make a difference. I still got the same performance.

Thanks and understood.

Enjoy the new rig and have a great day :-)

Kind regards

[ahsmatt7 1,199]

UPDATE!!!!

I forgot to mention things thst pictures can't show you.

 

When it comes to smoothness,

I found that the .cfg tweak I use at level 10 caused stuttering...a lot of it actually.

I found that setting fps to UNLIMITED with v-sync on and triple buffering gave the best experience.

[CaptainStraw 52]

19 minutes ago, ahsmatt7 said:

UPDATE!!!!

I forgot to mention things thst pictures can't show you.

 

When it comes to smoothness,

I found that the .cfg tweak I use at level 10 caused stuttering...a lot of it actually.

I found that setting fps to UNLIMITED with v-sync on and triple buffering gave the best experience.

What monitor are you using? Is it G-sync?

[ahsmatt7 1,199]

12 minutes ago, jacf182 said:

What monitor are you using? Is it G-sync?

No g-sync...just a Samsung monitor.

[gboz 157]

52 minutes ago, ahsmatt7 said:

as well as show those of you who have a 2xxxk i7 what the upgrade will bring if you do decide to upgrade to an 8700k.

Do you have any of your 2700K fps available for comparison?

gb.

[ahsmatt7 1,199]

7 minutes ago, gboz said:

Do you have any of your 2700K fps available for comparison?

gb.

I dont...i apologize. I only had one powers supply so I took it from my old computer and moved it to the new one.

I figured people with these older processors already had an idea of what performance they are already getting with their current rig.

[gboz 157]

24 minutes ago, ahsmatt7 said:

I dont...i apologize. I only had one powers supply so I took it from my old computer and moved it to the new one.

Ok no worries.

gb.

I notice you only use MSAA.
It would be very interesting if you could do a KDEN run again with say 4SSAA and see if DL still has no effect on fps.

Thanks,
gb.

[mtrainer 136]

Hi,

Thanks for posting.   Planning on building a similar rig with next spring's tax refund.  This year was devoted mostly to a Warthog HOTAS, LED keyboard, and the crosswinds rudder pedals.   Next step is the main rig, which is way overdue for an update. 

Mark

 

 

[MarkW 1,490]

Thanks for doing this.  Did you try a few scenarios with Dynamic Lights off?  I know you tested Dynamic Lights On and then had all lights off on the aircraft but wonder what the performance is like when you just use the Sim with Dynamic lights off and the aircraft lights on?

Also would be interested in what happens with some AI turned on.   If you have MT6 can you post some number with 10% traffic?

[RichieFly 725]

Thanks for this.

Enjoy your new rig.

[ahsmatt7 1,199]

25 minutes ago, MarkW said:

Thanks for doing this.  Did you try a few scenarios with Dynamic Lights off?  I know you tested Dynamic Lights On and then had all lights off on the aircraft but wonder what the performance is like when you just use the Sim with Dynamic lights off and the aircraft lights on?

Also would be interested in what happens with some AI turned on.   If you have MT6 can you post some number with 10% traffic?

I do have MT6....is it v4 compatible? I am quickly finding out how much of a pain it is to realize that not everything was made 64 bit compatible. I am VERY curious to see what AI traffic does.

 

Also, Ill run the same things with dynamic lights turned off in the sim. The only issue is that if dynamic lights are disabled, then PMDG airplanes wont have lights.....period. That's all I have installed right now. 

[Anxu00 176]

4 hours ago, ahsmatt7 said:

It seems as if the performance hit from the Dynamic lights depends on how the airport was developed???? Who knows. 

Yes it is.  If the airport is not built native with V4 SDK dynamic light performance dropped.  If you search the PMDG forum, Robert posted a topic on this specific issue.  Make sure you actually have version 2 of KFSO from FB.