Jump to content

Recommended Posts

http://forum.avsim.net/topic/329116-fsxmark11/?p=2417335

Its the 3770 and memspeed with loose timings

Based on those linked results it looks like you are both right, difference is due to 3770k vs 4770k architecture and slower memory timings. Interesting comparisons.

Thanks,

Ted


3770k@4.5 ghz, Noctua C12P CPU air cooler, Asus Z77, 2 x 4gb DDR3 Corsair 2200 mhz cl 9, EVGA 1080ti, Sony 55" 900E TV 3840 x 2160, Windows 7-64, FSX, P3dv3, P3dv4

Share this post


Link to post
Share on other sites

It's just disabled in nVidia control panel. I doubt removing the second card will make much difference though. The tested performance difference between pcie 3.0 16 x vs 8x is virtually negligible.

 

Is that tested performance difference in FSXmark11 or in general? I know that in general it doesn't make much of a difference.

However PCIe 2.0 x16 Vs x8 makes quite a bit of difference on a i7 860 @ 3.8 GHz in FSXmark11.

This is a graph plotting the FPS if you leave FSXmark11 running for 1200s (20 min). The actual FSXmark11 benchmark is the first 300s of this graph.

 

FSXmark11%20PCIe%202.0.jpg

 

The Haswell @ 4.5 GHz with faster RAM will have CPU grunt that's more than 50% faster than the Lynnfield @ 3.8GHz, so it wouldn't surprise me if PCIe 3.0 x8 will start to hold you back (same bandwidth as PCIe 2.0 x16). But as you can see, in real life it shouldn't be an issue if you lock your FPS to 30 as PCIe 2.0 x16 is capable of that.

Share this post


Link to post
Share on other sites

Since I've dug up a few old analysis of the FSXmark11 benchmark I've done. Here's a graph showing the effect of overclocking my GTX470 up to 810MHz from stock 607MHz (+33%) letting the FSXmark11 run for 1200s.

FSXmark11GfxOC810.jpg

 

During the first 300s that is the actual FSXmark11 benchmark there's only two short periods where the stock GTX470 is limiting Lynnfield @ 3.8 GHz slightly. Put in a 50% faster CPU and a faster GPU should make more of a difference, but as you can see, FSXmark11 is still mostly CPU limited.

 

Finally, here you can see why we discard the first run of the benchmark as it's giving a lot lower FPS:

FSXmark11%20Esc.jpg

 

It shows clearly how the FPS is a lot lower when FSX has to 'setup' the scene. When we exit the benchmark with Esc and re-run it we get a lot higher FPS during the time the aircraft flies over an area where it hasn't been before during that benchmark run. The scene has stayed in the GPUs memory giving us much higher FPS there. As you can see, almost the full 300s actual FSXmark11 part is very affected, while most of the time after that the aircraft just circles over the same area over and over again only drifting away over new areas slowly. If we were to close down FSX fully between the runs, the GPUs memory will be flushed and we'll end up with the lower FPS all the time. That's how FSX work in real life when we fly along enroute.

Share this post


Link to post
Share on other sites

Is that tested performance difference in FSXmark11 or in general? I know that in general it doesn't make much of a difference.

 

 

I was just talking in general terms. I know there will be some difference but should be barely noticable in general use as you said.

 

To update my findings, my board doesn't have the ability to disable my second graphics cards pcie 3.0 slot so I haven't tested running one card at x16 vs x8. I'll play around with that if I have time over the weekend.

 

I found my RAM had two different XMP profiles which oddly enough are identical to each other, but in the process of testing I ran FSXMark 11 again on both profiles with identical results.

 

Now the interesting part. Before the last run where I was getting about 48 average I had run Razer Gamebooster to close any programs etc. This time I didn't. Results are below.

 

Using Gamebooster:

 

 

Without Gamebooster:

 

 

 

I was able to replicate the results, so it wasn't just a one off. I'm not sure what Gamebooster was doing to hurt my results but needless to say I won't be using it anymore.

 

I've now hit the 50 FPS average I was looking for but I'm sure I can squeeze out a few more FPS if I learn how to tighten up my RAM timing a bit more and also bump up to 4.6Ghz (probably my limit without delid if I care about my temps when running stress tests).

 

Overall observation just watching the tests running. The increase in average FPS over my previous 4.9Ghz 2600K system is only about 3-4FPS but in terms of smoothness, to me it was noticibly smoother. It might just be a placebo, but I think it has more to do with the minimum FPS not dropping below 30.

Share this post


Link to post
Share on other sites

Imc handle High clocks better than IB but i not sure that my mems ar good for hi speed low latency. mange to run fsmark11 @2993 with loose timings ,

Shall test a pair MFR mems soon.

 

Take a lock at it , but its best to publish this before :

 

Mission completed

 

60.5-5.2.gif

That's one impressive result, congratulations!

 

What batch # is your 4770k?


Richard Åsberg

Share this post


Link to post
Share on other sites

Thanks

Batch 310B479

Not a good one i have test some other this one dont scale good on voltage .

Its hard to find a good one that scale well on voltage then if you shall run it on ln2 you

Have the coldbug from a bad -100C to a good -140C and that they scale well from 1.5 to 2.0 volt not stop scale at 1.7v

This stop scale at 1.65v and coldbug at -110C ( its only for the sim now)

 

A good chip shall do prime @5.0 ghz 1.3-1,35v with ht enabled on water.

 

The results was run early , its was not with high uncore

Have tuned it a bit more run the mems at 2400 with thigther timings cl8 and higher uncore.

 

With SS @5.1ghz uncore 4.8 mem 2400cl 8. Avg 61.5fps ( ht on)

Water @4.8 ghz uncore 4.5 mem 2400cl8 Avg 57.3fps ht off

 

If a find a good one a shall run the tests one more time, should be possible to do 60fps with watercooling.

Share this post


Link to post
Share on other sites

Found a L311B512 chip today and although it seems to be much of a lottery I've read lots of good things about this batch on many different sites.

 

Still haven't made up my mind though if I should go for the upgrade...sigh...


Richard Åsberg

Share this post


Link to post
Share on other sites

 

 


If a find a good one a shall run the tests one more time, should be possible to do 60fps with watercooling.

Watercooling can handle the heat but the voltages become almost logarithmic above 4.6 GHz.

 

HLJAMES

Share this post


Link to post
Share on other sites

Hi

Did a fast test with my spare 4670k not tuned the mems for max performance.

Cooling water loop 2x d5

 

My conclusion is that this CPU most be best Bang for bucks its faster than the IB 3770k.

Share this post


Link to post
Share on other sites

AMD Phenom X4 955 (overclocked to 3800), 8gb ram, geforce 760 (2gb)

Max. settings gives me 10 fps. Meduim (aircraft quality in priority) gives me 30 (not smooth at all). What am I doing wrong? Or it's just my phenom sucks?

Share this post


Link to post
Share on other sites

Nice one! First result over 60FPS in FSXmark11. I knew iit would be you Hasse=) Congrats

 

Yes Sir,

 

Well done!

 

Kind regards,

Share this post


Link to post
Share on other sites

Here are some runs with my C9-2400MHz ram at 2400MHz, 2000HMz and 1600MHz.

Further confirms that speed does matter with a 4770K-Z87 platform.

 

 

Hardware Configuration      

CPU: 4770K @ 4.6GHz / NB @ 4400MHz HT Off  

Ram: 2 * 4GB 2400MHz (9-11-11-31-1T)  

GPU: Gigabyte GV-N780OC GTX 780 @ 954MHz  

OC SSD: 840 Pro 256GB      

FSX SSD: 840 Pro 256GB                  

FSXMark11 Results        

Test     Frames    Time (ms)    Min         Max         Avg

2         15548       300000       32.000    66.000    51.827

3         15290       300000       32.000    65.000    50.967

4         15320       300000       31.000    66.000    51.067

AVG    15386       300000       31.667    65.667    51.287

 

 

Hardware Configuration      

CPU: 4770K @ 4.6GHz / NB @ 4400MHz HT Off  

Ram: 2 * 4GB 2000MHz (9-11-11-31-1T)  

GPU: Gigabyte GV-N780OC GTX 780 @ 954MHz  

OC SSD: 840 Pro 256GB      

FSX SSD: 840 Pro 256GB                  

FSXMark11 Results        

Test     Frames     Time (ms)     Min         Max         Avg

2         15179        300000        32.000   65.000    50.597

3         14930        300000        32.000   65.000    49.767

4         14655        300000        31.000   65.000    48.850

AVG    14921.33   300000        31.667   65.000    49.738

 

 

Hardware Configuration      

CPU: 4770K @ 4.6GHz / NB @ 4400MHz HT Off  

Ram: 2 * 4GB 1600MHz (9-11-11-31-1T)  

GPU: Gigabyte GV-N780OC GTX 780 @ 954MHz  

OC SSD: 840 Pro 256GB      

FSX SSD: 840 Pro 256GB                  

FSXMark11 Results        

Test      Frames       Time (ms)     Min         Max       Avg

2          13934          300000        29.000   61.000   46.447

3          13790          300000        29.000   62.000   45.967

4          13943          300000        30.000   61.000   46.477

AVG     13889          300000        29.333   61.333   46.297

________________________________


For some reason the forum will not let me make two separate posts, it keeps adding the second post to the first one.  Pretty strange!

 

Anyway here is a run from my current STZ87-4770K.  This is 2 FPS faster than the last STZ87-4770K I had that I used to make the different ram speed runs posted here.  Both setups used the same GTX 780, ram and Win7 / FSX install.  Since I did not change the motherboard type I did not need to install Win7 from scratch when I changed the MB and CPU.

 

 

Hardware Configuration      

CPU: 4770K @ 4.6GHz / NB @ 4400MHz HT Off  

Ram: 2 * 4GB 2400MHz (9-11-11-31-1T)  

GPU: Gigabyte GV-N780OC GTX 780 @ 954MHz  

OC SSD: 840 Pro 256GB      

FSX SSD: 840 Pro 256GB                  

FSXMark11 Results        

Test    Frames     Time (ms)      Min          Max           Avg

2        16057        300000         31.000     72.000      53.523

3        15863        300000         33.000     72.000      52.877

4        16059        300000         32.000     73.000      53.530

AVG   15993        300000         32.000     72.333      53.31

 

If I have not added any addons to FSX by the time I OC to 4.7GHz I will post the results for a 4.7GHz 4770K OC.

Share this post


Link to post
Share on other sites

4.7GHz results.

 

Hardware Configuration       CPU: 4770K @ 4.7GHz / NB @ 4400MHz HT Off

Ram: 2 * 4GB 2400MHz (9-11-11-31-1T)

CPU: Gigabyte GV-N780OC GTX 780 @ 954MHz

OC SSD: 840 Pro 256GB

FSX SSD: 840 Pro 256GB

FSXMark11 Results

Test   Frames     Time (ms)     Min         Max        Avg

2       16423        300000        33.000   75.000    54.743

3       16195        300000        32.000   74.000    53.983

4       16246        300000        33.000   74.000    54.153

AVG  16288        300000        32.667   74.333    54.293

Share this post


Link to post
Share on other sites

Finally finished my new build, and am surprised by the results on Haswell.

 

Hardware Configuration:      

CPU: Intel i5 4670k @ Stock

RAM: 2x4GB 2666 (11-13-13-35-2T)

GPU: MSI GTX670 2gb Power Edition

HDD: Seagate Hybrid 1TB            

 

FSXMark11 Results:      

Test Frames  Time (ms)  Min  Max  Avg

   2 13445        300000 27.0 62.0 44.8

   3 13238        300000 27.0 62.0 44.1

   4 13139        300000 27.0 61.0 43.8

AVG 13274      300000 27.0 61.7 44.2

 

Will now clock and post back


P3D v4.5 MSFS2020 Hisense 50" 4K TV

Ryzen 5800X, 32gb DDR 3600mhz, MSI B550 PRO VDH WiFi, MSI 6900XT Z Trio, Gammaxx L360, 1TB NVMe Boot/FS2020 Drive, 1TB NVMe P3D Drive, 1Tb Crucial SSD Storage Drive, Saitek Yoke, Pedals, Radio Panel, Switch Panel, 2 x FiPs

UKV6427

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...