Page 2 of 18 FirstFirst 123456789101112 ... LastLast
Results 21 to 40 of 360

Thread: The TWC Total War Benchmark thread - compare your real TW performance

  1. #21
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Good morning everyone, there are new submitted results and first conclusions (post #7)

    I hope that your engagement to participate will not drop! We need more participants with mid and entry rigs, older dual core and quad CPUs Intel E / Q Series.
    If you do not have a DX11 capable graphics card you can at least submit your CPU benchmark.
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  2. #22

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    [QUOTE=I hope that your engagement to participate will not drop! We need more participants with mid and entry rigs, older dual core and quad CPUs Intel E / Q Series.
    [/QUOTE]

    Planning on posting today or tomorrow with my y500 with 750msli. More curious than anything.

  3. #23
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Thanks! Looking forward!
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  4. #24
    babydoc's Avatar Semisalis
    Join Date
    Mar 2009
    Location
    In a big house.
    Posts
    490

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    I think that some members are "scared" of whats expected of them to do theses tests.
    It takes some time and commitment.
    And the path in the guide must be a little different on different comps dont you think?
    But I agree that we need a lot more comps here....

    C:\Users\Babydoc\AppData\Roaming\The Creative Assembly\Shogun2\scripts vs.

    %appdata%\the creative assembly\shogun 2\scripts (<----from the guide.)

  5. #25
    SamueleD's Avatar Campidoctor
    Join Date
    Nov 2012
    Location
    Switzerland
    Posts
    1,555

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by babydoc View Post
    I think that some members are "scared" of whats expected of them to do theses tests.
    It takes some time and commitment.
    And the path in the guide must be a little different on different comps dont you think?
    But I agree that we need a lot more comps here....

    C:\Users\Babydoc\AppData\Roaming\The Creative Assembly\Shogun2\scripts vs.

    %appdata%\the creative assembly\shogun 2\scripts (<----from the guide.)
    There's also the fact that AppData is a hidden folder and must be made visible if you want to enter by clicking on it rather than put the address in, which is how most people do it.

  6. #26
    babydoc's Avatar Semisalis
    Join Date
    Mar 2009
    Location
    In a big house.
    Posts
    490

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by SamueleD View Post
    There's also the fact that AppData is a hidden folder and must be made visible if you want to enter by clicking on it rather than put the address in, which is how most people do it.
    Good eye dear ,Sir.
    I forgott about that.

    Maybe a programmer good make an .exe or .bat for this?

  7. #27
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Using this %appdata% in the explorer leads you directly into your directory no matter it is hidden.

    I have to commit that it is not that easy than I thought to get this working for tech newbies but the problem is the dx 11 high test does not use all ultra settings and options additionally he is downgrading on some gpus.
    Last edited by alQamar; August 08, 2013 at 10:10 AM.
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  8. #28
    SamueleD's Avatar Campidoctor
    Join Date
    Nov 2012
    Location
    Switzerland
    Posts
    1,555

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by alQamar View Post
    I don't see a difference on my mobile
    There is a space between "shogun" and "2" on some of the addresses you posted, if one just copy pasted them (like lazy me ) they wouldn't work, because in the correct address it's "shogun2" without a space.

  9. #29
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Ok fixed that thanks
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  10. #30

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Ok so where is the info mate showing that TW uses all four cores? Last time i checked its two cores 90% two cores 20% thats not properly four core optimization.Multicore would be all four cpus equal load around 70%.


    Also a few mistakes i noticed in here.

    The difference is as follows:
    145 FPS with an hexa (6) core i7 and 2x TITAN SLI with hyperthreading enabled and modified 32 threads setting in the preferences.script.txt (refer post #4)
    152 FPS with an hexa (6) core i7 and 2x TITAN SLI with hyperthreading disabled and modified 32 threads setting in the preferences.script.txt (refer post #4)

    This is due to negative scaling and it is not just Totalwar that needs HT disabled on the six cores but Battlefield 3 as well.Basically its getting overloaded and it gets much better linear scaling with HT off.And why is the settings at 32 threads? The 6 core is six threads only and that turns into 12 with HT enabled.Maybe if he tried with HT OFF and experimented with 6-12 threads he might get better fps.


    Anisotropic Filtering: 16x
    Ambient Occlusion : Quality (leave standard setting on entry or laptop graphics)
    Anisotropic filtering: Application-controlled
    Antialiasing - FXAA: On (only set this on program related settings, as some Windows Applications will get blurry (Office 2013, Messenger, Firefox)
    Antialiasing - Gamma correction: On
    Antialiasing - Mode: Application-controlled
    Antialiasing - Setting: Application-controlled
    Antialiasing - Transparency: 4x / 8x Supersampling (leave it disabled if you have an entry card or laptop. Test this settings and lower or turn off if it causes low FPS. )
    CUDA - GPUs: All
    Maximum pre-rendered frames: leave standard settings
    Multi-display/mixed-GPU acceleration: Single display performance mode (if you have a single monitor)
    Power management mode: Adaptive
    Texture filtering - Anisotropic sample optimization: Off
    Texture Filtering Quality: high quality (leave standard setting on entry or laptop graphics)
    Texture filtering - Negative LOD bias: Clamp
    Texture filtering - Trilinear oprimization: On
    Threaded optimization: Auto
    Tripple buffering: On
    Vertical sync: Adaptive (This offers a brilliant mix of non tearing and performance)
    Triple buffering and gamma correction are OpenGL settings so have no effect on Totalwar.And power management should be best performance over adaptive as adaptive can mess with overclocks.I use these settings but this is for Quality over Performance and will lose a few fps in favour of sharper better graphics.



    Ambient Occlusion : Quality
    Anisotropic filtering: Application-controlled
    Antialiasing - FXAA: Off
    Antialiasing - Gamma correction: Off
    Antialiasing - Application-controlled
    Antialiasing - Transparency: Multisampling
    CUDA - GPUs: All
    Maximum pre-rendered frames: 1 (change to 3 if frame dropping occurs due to weak CPU)
    Multi-display/mixed-GPU acceleration: Single display performance mode (if you have a single monitor)
    Power management mode: Prefer Maximum Performance
    Texture filtering - Anisotropic sample optimization: Off
    Texture Filtering Quality: high quality
    Texture filtering - Negative LOD bias: Clamp
    Texture filtering - Trilinear oprimization: OFF
    Threaded optimization: On
    Tripple buffering: Off
    Vertical sync: Application controlled

    This here gives highest possible image quality on Nvidia cards.It disables all the nasty fps saving graphics killing settings and makes everything set up for multisampling antialiasing and anisotripic filtering to be controlled ingame.I use 16x AF 4xMSAA with these settings.
    Last edited by Jedi; August 08, 2013 at 03:43 PM.
    3570K 4700mhz cooled with Corsair H80 // Asus Z77 // MSI GTX 580
    16GB 2400mhz DDR3 // Crucial M4 256GB Raid 0 // Dell 2007WFP

  11. #31
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by Jedi View Post
    Ok so where is the info mate showing that TW uses all four cores? Last time i checked its two cores 90% two cores 20% thats not properly four core optimization.Multicore would be all four cpus equal load around 70%.


    Also a few mistakes i noticed in here.




    This is due to negative scaling and it is not just Totalwar that needs HT disabled on the six cores but Battlefield 3 as well.Basically its getting overloaded and it gets much better linear scaling with HT off.And why is the settings at 32 threads? The 6 core is six threads only and that turns into 12 with HT enabled.Maybe if he tried with HT OFF and experimented with 6-12 threads he might get better fps.




    Triple buffering and gamma correction are OpenGL settings so have no effect on Totalwar.And power management should be best performance over adaptive as adaptive can mess with overclocks.I use these settings but this is for Quality over Performance and will lose a few fps in favour of sharper better graphics.



    Ambient Occlusion : Quality
    Anisotropic filtering: Application-controlled
    Antialiasing - FXAA: Off
    Antialiasing - Gamma correction: Off
    Antialiasing - Application-controlled
    Antialiasing - Transparency: Multisampling
    CUDA - GPUs: All
    Maximum pre-rendered frames: 1 (change to 3 if frame dropping occurs due to weak CPU)
    Multi-display/mixed-GPU acceleration: Single display performance mode (if you have a single monitor)
    Power management mode: Prefer Maximum Performance
    Texture filtering - Anisotropic sample optimization: Off
    Texture Filtering Quality: high quality
    Texture filtering - Negative LOD bias: Clamp
    Texture filtering - Trilinear oprimization: OFF
    Threaded optimization: On
    Tripple buffering: Off
    Vertical sync: Application controlled

    This here gives highest possible image quality on Nvidia cards.It disables all the nasty fps saving graphics killing settings and makes everything set up for multisampling antialiasing and anisotripic filtering to be controlled ingame.I use 16x AF 4xMSAA with these settings.

    Hi Jedi. I am open for suggestions, but there are no mistakes at least not those you mentioned.

    Next time please use PM respecting the thread rules.
    But now as your views are posted I have to comment them to avoid irritations.

    Quad core support
    I ran several hours of benchmarking and released my results in the multicore thread (page 7)

    Running my cpu with less than 4 cores affected the performance measurably. That's why I can proove it is using 4 cores and 4 cores are more performant than 3 or 2. I never said TW is using all 4 cores efficiently. your statement when multicore supports begin is not objective here.


    Hyperthreading conclusion:
    First of all this is about TW not other games. Civ5 uses hyperthreading par excellence for example.
    There was nothing wrong to say an i7 is not worth to have for Total War.
    You added another example but that's off topic.

    A general disabling of hyperthreading is a bad suggestion! You will loose performance on an i3 playing TW! But gain performance if you have it enabled and use the correct thread tweak.
    I give very detailed information here that are well tested and not grabbed elsewhere.
    I hope you can respect this and not join and throw around false positives.

    Number of threads
    This setting has nothing to do somehow with the number of physical or virtual threads of your cpu but the threading relays to number of threads used by the total war engine. Boxleitnerb tested a lot of settings to find the sweet spot for an i7 hexa core and baby doc helped the same to find it for the AMD octa core.

    Just believe we are doing a professional job. While you are currently guessing around that the number of threads settings should (could) be wrong. I don't like this either getting blamed or at least accused for mistakes with no proove mate.

    Nvidia settings
    Double and triple buffering is used since DirectX7 and is also used by openGL. The triple buffering is needed for better Vsync results.

    Furthermore you left Vsync on "application controlled" that's wrong as application will not likely use adaptive Vsync that offers best performance and non tearing as you spoke about having the best settings used for performance and quality.


    Adaptive performance is a perfect setting for the masses. This thread is for the masses as well. It may be you are right in particular with overclocking but all vendor oc cards run eco (power efficient and silent) and fast with adaptive performance settings. Especially paired with adaptive Vsync that performance setting is not useful as the fps capping with adaptive Vsync and adaptive power management saves a lot of overhead calculations and so save energy and heat aswell cooling noise then.


    Next time please pm me instead of postulating stuff just to find something wrong.

    However you are welcome to send your benchmarks in.
    Please also notice that the driver suggestions should not be applied prior the benchmarks as I cannot validate the correct settings via the log files. That's why the users should use the "average" standard driver settings, so we can expect an objective comparison.
    Last edited by alQamar; August 09, 2013 at 05:34 PM.
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  12. #32
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Hello everyone. I am happy this thread got already more than 2000 views but only 10 submissions have been done yet.

    I would like to know why there are so less replies and submissions.

    Are there questions?

    Anything not working or to hard to do understand?

    Happy to get any feedback so I can improve the instructions.
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  13. #33

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Interesting article I just found, it shows CPU benchmarks for shogun 2 stock and if you OC. For example if you run stock Intel 2500k 3.3ghz you get min 25FPS while average is 28fps. If you OC all the way up to 5ghz, you get 33fps min and 40fps average. Not much of an increase at all for crazy OCing. Warscape is pure trash....

    http://www.bit-tech.net/hardware/201...l-cpu-review/5
    Last edited by Jasheon; August 10, 2013 at 10:31 PM.

  14. #34

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Testing with 4 Threads:
    11_08_05_57_59_benchmark1.log.txt 11_08_06_01_25_benchmark1.log.txt

    5_57_59 is GPU benchmark: GTX 780 SC OCX - 108.8 fps (SuperClocked OCX from EVGA runs at 967MHz GPU Clock with 1020MHz in Boost-Mode and 1502MHz RAM frequency)
    6_01_25 is CPU benchmark: i7 4770K - 45.5 fps (the i7 4770K in my PC is overclocked to run at 4500 MHz via a Bus Speed of 125 MHz)

    Testing with 16 Threads:
    11_08_20_21_46_benchmark1.log.txt CPU-Z Log.txt 11_08_20_15_19_benchmark1.log.txt

    20_15_19 is GPU benchmark: GTX 780 SC OCX - 108.7 fps
    20_21_46 is CPU benchmark: i7 4770K - 52.2 fps

    Testing with 16 Threads @ 1666MHz downclocked RAM:
    12_08_11_29_11_benchmark1.log.txt 12_08_11_32_09_benchmark1.log.txt CPU-Z Report 2.txt

    11_29_11 is GPU benchmark: GTX 780 SC OCX - 108.3 fps
    11_32_09 is CPU benchmark: i7 4770K - 48.5 fps

    Cool to see that the RAM clock is giving me a 7,6% boost in this benchmark. I wasn't sure how useful it would exactly be to buy such high-end ram for games.

    Rest of my setup for comparability (mobo and ram might matter, ssd too?):

    MoBo : Asus Maximus VI Hero ( Z87 Board )
    RAM : CMD16GX3M4A2666C11 ( That's 16GB of Corsair DDR3 RAM overclocked at 2666MHz )
    OS : Win7 Pro
    SSD : Samsung 840 Pro series 2,5" 512 GB ( I put Shogun 2 on the SSD )

    The OS is still very clean since I build this PC three days ago for the express purpose of running Rome II on max settings. So I like that I can show it off a bit here if it's ok to gloat a bit here
    Last edited by Dominian; August 12, 2013 at 05:39 AM.
    My signature is running from the battlefield. A SHAMEFUR DISPRAY!

  15. #35
    crzyrndm's Avatar Artifex
    Join Date
    Feb 2011
    Location
    New Zealand
    Posts
    2,576

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by Jasheon View Post
    Interesting article I just found, it shows CPU benchmarks for shogun 2 stock and if you OC. For example if you run stock Intel 2500k 3.3ghz you get min 25FPS while average is 28fps. If you OC all the way up to 5ghz, you get 33fps min and 40fps average. Not much of an increase at all for crazy OCing. Warscape is pure trash....
    Er...

    5GHz is a 51% increase over 3.3GHz (EDIT only 35% over the 3.7GHz turbo boost of the stock 2500)
    40fps is a 43% increase over 28fps

    Don't see your problem...

    EDIT

    It has just been pointed out to me that the 2500k will turbo boost up to 3.7GHz putting the 5GHz clock only 35% higher making the fps gain greater than 1:1 for clock speed.
    Last edited by crzyrndm; August 11, 2013 at 06:00 AM.
    It’s better to excite some and offend others than be bland and acceptable to all
    Creating a mod.pack with PFM - Database Table Fragments

  16. #36
    alQamar's Avatar Citizen
    Join Date
    Jul 2012
    Location
    Dortmund, Germany
    Posts
    5,963

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by Dominian View Post
    11_08_05_57_59_benchmark1.log.txt11_08_06_01_25_benchmark1.log.txt

    5_57_59 is GPU benchmark: GTX 780 - 108.8 fps
    6_01_25 is CPU benchmark: i7 4770K - 45.5 fps (the i7 4770K in my PC is overclocked to run at 4500 MHz via a Bus Speed of 125 MHz)

    Rest of my setup for comparability (mobo and ram might matter, ssd too?):

    MoBo : Asus Maximus VI Hero ( Z87 Board )
    RAM : CMD16GX3M4A2666C11 ( That's 16GB of Corsair DDR3 RAM overclocked at 2666MHz )
    OS : Win7 Pro
    SSD : Samsung 840 Pro series 2,5" 512 GB ( I put Shogun 2 on the SSD )

    The OS is still very clean since I build this PC three days ago for the express purpose of running Rome II on max settings. So I like that I can show it off a bit here if it's ok to gloat a bit here
    No I do only need all logs cpu-z log is missing at moment please edit your posting and attach it please.

    I downloaded your logs now and noticed you set the number of threads to 4. This is wrong for your CPU. You have an i7 so please set this to 16 and re run both benchmarks again and submit the logfiles and CPU-Z as said. Sorry for bothering you with that but they are not valid for submission at the moment.
    Last edited by alQamar; August 11, 2013 at 05:27 AM.
    NEW: Total War Saga: Britannia benchmark thread - last update: 10.05.2018
    HOW-TO-step-up-from-MBR-CSM-LEGACY-BOOT-to-UEFI-GPT
    Many of my past contributions in the time from 2011-2017 will contain content that now show broken links. Unfortunately I had to delete all pictures linked on TWC that were hosted on imageshack.us. Read why
    If you are missing anything of interest, please let me know. Sorry for any inconvinience caused.

  17. #37

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Hey alqamar an AMD processor A6 dual 2.1-2.6 GHz will run total war? I think yes because is better than i 3 intel and it is like i5

  18. #38
    SamueleD's Avatar Campidoctor
    Join Date
    Nov 2012
    Location
    Switzerland
    Posts
    1,555

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by Aragorn-Elessar View Post
    Hey alqamar an AMD processor A6 dual 2.1-2.6 GHz will run total war? I think yes because is better than i 3 intel and it is like i5
    Sorry dude, this thread is for benchmark related questions only, you should post this on the hardware advice thread here, we'll all be happy to answer.

  19. #39
    DramaBelli's Avatar Ministry of Silly Walks
    Join Date
    Feb 2010
    Location
    Rome
    Posts
    3,816

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    hi there, benchmark submitted via mail..hope everything is ok! thanks

  20. #40
    Biggus Splenus's Avatar Primicerius
    Join Date
    Mar 2012
    Location
    South Australia
    Posts
    3,547

    Default Re: The TWC Total War Benchmark thread - compare your real TW performance

    Quote Originally Posted by DramaBelli View Post
    hi there, benchmark submitted via mail..hope everything is ok! thanks
    Nice to see you've finished your build
    | R5 3600, RTX 2060, MSI B450I, 32GB 3200MHz CL16 DDR4, AX760i, NH-U12S |

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •