Not signed in (Sign In)

Discussion Tag Cloud

Categories

Vanilla 1.1.9 is a product of Lussumo. More Information: Documentation, Community Support.

  1.  

    No Limits Coaster 2

    BENCHMARKING

    How to get your results:

    Step 1:

    Go to your desktop and find the shortcut for NoLimits 2. Once you have found it, right click it.


    Step 2:

    Choose "properties", which will open a small window. In this window choose the "Shortcut" tab which usually open automatically.


    Step 3:

    Find the "Target:" path. At the very end of the path address, add the parameter -t.


    Step 4:

    Click apply and double click on your icon to run NoLimits 2. This will now launch the program in "timedemo mode". Once open, click "Play" which will bring a "Select park..." window up, choose the "Library" tab, and then select "Forest Mountain Park", which is the official NLE Benchmarking park.


    Step 5:

    After selecting or double clicking, the park will load and start the ride automatically.


    Once loaded, you will notice you cannot use your mouse (some keyboard functions will still work, but do not press any as this will render your results invalid), and that there are some new words and numbers at the bottom of your screen.


    Do not do anything as the benchmarking is now under way. Resist the temptation to speed up the lift or do anything else on your computer! If things are going to plan, you should reach 69% just after the tunnel.


    Step 6:


    You'll pull up in the brakes, wait for the train to leave, pull into the station and still have a few percent to go. Let the benchmarking run its course. Ready your fingers on the windows and PrtScn buttons, then when the test gets to 100% take a screenshot.


    These are your results, post them here for all to see!
  2.  
    EDIT: nvermind
    • CommentAuthorole
    • CommentTimeJul 16th 2014 edited
     
    Thank you very much for the great explanation. You can also take the Total number of frames for the performance indication. The time demo runs for 180 seconds (3 minutes) exactly, so the average fps is identical to Total number of frames divided by 180. The total number of frames is stored in the timedemo.txt log file, so you can leave the room, fix something to eat or visit the bathroom while running a time demo and then later check the results of the time demo by looking at the log file. The log files are stored somewhere in C:\Users\(myname)\AppData\Roaming\com.nolimitscoaster.nolimits2. The AppData folder is typically a hidden folder, so you first need to make hidden folders visible.
  3.  

    My benchmarking on an Alienware 17

  4.  
    I got an average of 47.19fps

    Specs:
    AMD 6300 Hex Core Processor
    8GB ram
    Nvidia Geforce GTX 770

    Might note it was running on 3 screens.
  5.  
    Custom built workbench (not overclocked, yet):
    i7 extreme 4930k
    32gb 2133hz ram
    2x GTX 780 sli

    Note: Strange thing happened, was running at an average 110fps the whole ride until the train pulled into the station, then the frames started dramatically decreasing.
    •  
      CommentAuthordapalm
    • CommentTimeJul 16th 2014
     
    ^ Same here (well, from 48 to 41FPS).

    I7 4770
    16Gb 1600mhz RAM
    Nvidia GT640

    I think everyone should post their computer specs.
    •  
      CommentAuthorProjektion
    • CommentTimeJul 16th 2014
     


    i5 3570
    8GB RAM
    Nvidia GTX 770
  6.  
    52.30 avergae fps

    I too had higher fps until stopping in the station.
    • CommentAuthor1amediacbus
    • CommentTimeJul 16th 2014 edited
     
    Forest Mountain Park: 175.58 FPS Average. Max Graphics.

    Tango Park: 119.21 FPS Average. Max Graphics.
    •  
      CommentAuthorBolt789
    • CommentTimeJul 16th 2014
     
    Is it normal for it to lag more while benchmarking than when not? That's what happened to me.



    AMD A10-6800K
    8GB RAM
    NVIDIA GeForce GTX 750 Ti
  7.  
    Before we get too deep into this thread, I think we need to standardize a couple of things to make sure everyone is on the same page. When I first ran my benchmark, I had "use current time" enabled, which ran the sim at an in-game time of about 8:15 pm (the time in Greenwich, England, the default location setting in NL2). I ran it again, this time with "use current time" disabled and using a 60 minute day-night cycle and got very significantly different results. So, to help keep everyone's benchmarks as directly comparable with each other as possible, I say that everyone should use the following settings when benchmarking:

    1. Location: 54.48°N, 0.00°E (I don't know if this affects anything outside of the use current time function, but it's better to standardize it in case it does.)
    2. Day-Night Cycle: 60 Minutes
    3. All detail settings set to maximum
    4. Vsync disabled (IMPORTANT: make sure that it is also not enabled through the nVidia Control Panel or Catalyst Control Center)
    5. Triple buffering enabled (Again, not sure if it will affect anything greatly, but it's good to have a final say on what it should be set to.)

    Additionally, all of the following information should be provided:

    1. CPU, include it's clock speed if you've overclocked it. Eg. AMD FX-4300@4.4GHz
    2. Amount/type of RAM and its speed. Eg. 8GB DDR3 1600MHz
    3. GPU Eg. EVGA Superclocked ACX GTX 780

    And finally, I recommend that everyone everyone include a test run at 1920x1080. If you have a monitor natively uses a resolution other than 1080p, feel free to include a test run at native resolution, but be sure to include a 1080p test as well. This is simple to set up, if your monitor's resolution is greater than 1080p, it is as simple as selecting the 1920x1080 selection from the "fullscreen mode" dropdown menu in settings and running a benchmark. If your monitor runs at less than 1080p natively, you can force it to run at 1080p through the nVidia control panel or Catalyst Control Center. However, keep in mind that this will not magically turn your monitor into an actual 1080p monitor, it will only trick your computer into thinking it is and allow you to force your hardware to actually render everything at 1080p. The end result that you see will be downsampled to fit your monitor, but you will still see the same performance as you would while using a true 1080p monitor. This will not harm your monitor in any way. This is actually a well known method of antialiasing called OGSSAA. Learn more about it here if you are curious.

    So, all of that being said, here are my results:

    CPU: AMD FX-4300@4.4GHz
    RAM: 8GB DDR3 1600MHz
    GPU: EVGA Superclocked ACX GTX780

    This is a test run at my monitor's native resolution of 2560x1440:


    And this is a test forced at 1080p:


    Also, Ole, would it be possible to add min and max FPS counters to the time demo? I feel like that would also help to provide useful information for benchmarking.

    And lastly, can anyone else figure out why my 1080p run ended up so similar to my 1440p run? That is baffling me right now.
    • CommentAuthorole
    • CommentTimeJul 16th 2014 edited
     
    Good point. The sun's position on the horizon has a large effect on the performance, because it will affect the shadows. Typically, the lower the sun, the larger the impact, because the shadows will be flatter and the shadow volumes will be larger, so more objects will be taking into account for computing the shadow buffers. It is good to standardize on the time and location settings.

    Also, I recommend to turn off Depth-Of-Field effect in the view panel (F7). Depth of field will have a large impact on performance. I guess that it will typically only be used for creating some special screenshots, and have it off otherwise.

    @Vekoma Fan Boy: This park obviously is not fillrate limited on your computer. If changing screen resolution is not having a large impact on performance, other components than the GPU are most likely the bottleneck. This park has a lot of smaller objects (lots of supports and trees), that means lots of draw calls, so I am pretty sure this park is limited by D3D API, driver, RAM speed, or CPU speed.

    I can try adding more statistics (min/max fps) in upcoming versions. Adding min fps will be tricky, because NL2 is not doing any pre-caching before rendering the first frame when the park loaded. Without pre-caching, the first frame that will be rendered will be very slow because all the buffers need to load into the GPU for the first draw call and that will most likely make it the minimum fps.
  8.  
    Posted By: ole
    @Vekoma Fan Boy: This park obviously is not fillrate limited on your computer. If changing screen resolution is not having a large impact on performance, other components than the GPU are most likely the bottleneck. This park has a lot of smaller objects (lots of supports and trees), that means lots of draw calls, so I am pretty sure this park is limited by D3D API, driver, RAM speed, or CPU speed.


    If that's the case, then I strongly suspect that it is my CPU since it is relatively underpowered compared to my GPU. I will have to give the benchmark another shot once I get my 4790k and compare the results.

    Also, this is just a thought, if the first frame rendered is almost guaranteed to to be the slowest by a large margin, would it be at all possible to simply exclude that single frame from the minimum FPS count? I.e. simply use the second-longest frame time to calculate the effective minimum FPS? Obviously I'm not a game designer, but I'm just curious and throwing out some ideas.
    • CommentAuthorole
    • CommentTimeJul 16th 2014
     
    @Vekoma Fan Boy: Yeah, I guess tossing away the first frames for computing the lowest fps should do the trick. My computer has a slower GPU and faster CPU and I am getting a higher fps than you, so changing the CPU should make things faster on your computer. I do not see a big difference when changing resolution, too.

    My computer:

    CPU: Intel Core i5 4570 (3.2GHz quad core Haswell)
    RAM: 8GB DDR3 1600MHz
    GPU: EVGA GTX 750 Ti FTW

    Average FPS: 67.26
    •  
      CommentAuthorMr. Laptop
    • CommentTimeJul 16th 2014
     
    Okay. I just did a test using max NL2 graphic settings. I think the rest are untouched. I did some major computer tuneup, and below are specs and the final result.
    First my computer specs:

    Windows 8.0 64-bit OS
    AMD A6-4400M APU with Radeon HD graphics 2.70 GHz
    4.00 GB (3.46 usable) RAM.

    Forest Mountain park with I think zero interruptions, copied from the log file: Frames: 2234
    •  
      CommentAuthorstejorin
    • CommentTimeJul 17th 2014
     
    All detail settings in the middle, Advanced settings off
    Anti-Aliasing : Quality 2
    FOV: Default
    resolution:1680x1050
    gpu: Nvidia Pny Geforce GT 630
    RAM:4gbs
    cpu:Intel Pentium dual-core e5400 @2.70GHz
    nolimits2app 2014-07-16 16-09-55-562
  9.  
    Nobody here is getting as good of results as me. Lol. XD
    • CommentAuthorSpike
    • CommentTimeJul 17th 2014
     
    Alright, well I thought my computer was on the lower end for running NL2, but it seems to be little better than I expected comparing results with everyone else.


    I'm running NL2 through Parallels on a Macbook Pro.

    CPU: 2.3 GHz Intel Core i7

    RAM: 16 GB 1600 MHz DDR3

    GPU: NVIDIA GeForce GT 750M 2048 MB


    I got an average of 43.11 fps while benchmarking.
    • CommentAuthorole
    • CommentTimeJul 17th 2014
     
    @1amediacbus: Could you please provide a screenshot to proof the good results? No offense, but I highly doubt you are getting these good results.
  10.  
    I'll do it when I get back from my KD trip. I'm headed there tomorrow for a week, to see my family and to hit up KD. Ittl be a week or two before I can post my screenshot. I currently do not have my super computer with me ATM.
  11.  
    ^Isn't that convenient ;)
    •  
      CommentAuthoraw555000
    • CommentTimeJul 17th 2014 edited
     
    ^^ For 1 second I don't believe your getting that FPS unless you work at NASA and somehow installed it on a communications server with 192GB of RAM and like 900 Terabyte HDD's or work at Oxford University Science & Technology department and installed it on the new biological computer system they've invented with a processor that works in Terahertz... People with good computers aren't even getting half of your alleged FPS so I'm calling BS, I will retract that should legitimate proof be provided and that the proof has no evidence of tampering (I can spot Photoshop fakes a mile off), I'd also like to know your computer specs, perhaps with additional proof taken from the Microsoft system information screen?

    @Ole I'd be interested in knowing your FPS and your computer system.
    •  
      CommentAuthorXpress
    • CommentTimeJul 17th 2014
     


    Yes, well... Nobody ever said my PC was stellar...

    Dell Inspiron i537S

    Intel Core Duo 3.00GHz

    4GB RAM

    ATI RADEON 4300/4500 series
    •  
      CommentAuthorride_op
    • CommentTimeJul 17th 2014 edited
     
    @aw555000 Scroll up a bit for Ole's system and results.

    The drop in frames once you return to the station may have to do with the other 2 Timberliner trains being visible. Timberliner trains have a lot of detail and having both trains on screen may have an effect as you return to the station.

    -Ride_Op
    • CommentAuthornesdude
    • CommentTimeJul 17th 2014
     


    Intel i7 2600k - stock 3.4ghz
    g.skill 16GB RAM - 1600mhz
    EVGA 780 SC ACX (03G-P4-2784-KR), no additional OC
    Samsung 830 128GB SSD

    The_Architect, can you run your test again without SLI (with just one of the 780TI's)? I'm curious to see how many frames, if any, a second card brings to NL. I notice your RAM speed is very fast, and ole mentioned something about how RAM speed might actually make a difference. Everything in that machine is beast, though.

    I'm tempted to mess with some light overclocking to see if it helps. The only reason I don't OC is because I value a cooler work environment over additional speed.
    • CommentAuthorRazevil
    • CommentTimeJul 17th 2014
     


    Intel i7 960 3.2GHz
    12GB Ram
    AMD Radeon R9 270x
    • CommentAuthorspinba11
    • CommentTimeJul 17th 2014
     
    I'm not sure how to post photos but I got an average of 89 FPS.
    •  
      CommentAuthorUncleArly
    • CommentTimeJul 17th 2014
     
    I'm getting like an average of 6 FPS.


    FML.
    • CommentAuthorForceWorks
    • CommentTimeJul 17th 2014
     
    Ouch....

    Time run finished waaaaaaaaaaay after the tunnel on here...

    Dell Optiplex 790 Core i7-2600 CPU @3.40 GHz
    8.00 GB RAM DDR3
    Windows 7 64 Bit
    Asus nVidia GeForce GT210 1 GB DDR2 DVI/HDMI PCIE 2.0 Video Card EN210 SILENT/DI/1GD2(LP)

    My Graphics card is DEFINITELY the limiter here (its shite...)
  12.  
    Posted By: Vekoma Fan BoyAnd finally, I recommend that everyoneeveryone include a test run at 1920x1080.


    What if your comp's resolution cant go up to 1080p?
    •  
      CommentAuthorviking86
    • CommentTimeJul 17th 2014
     
    AMD Athlon(tm) 64 X2 Dual Core Processor 6000+ 3.00 GHz
    4 GB RAM (3.06 GB usable)
    NVIDIA GeForce GTX 560

    Image and video hosting by TinyPic
    •  
      CommentAuthordapalm
    • CommentTimeJul 17th 2014
     
    Made a test with the settings provided by Vekoma Fan Boy. Again, computer specs:

    I7 4770
    16Gb DDR3 1600mhz
    Nvidia GT640*
    1366x768

    Benchmark1



    * Default GPU settings

    Benchmark2





    * Max GPU settings.

    Was thinking to upgrade the GPU to a GTX760, but seeing how the 750 performs I may save some money.
    •  
      CommentAuthorstejorin
    • CommentTimeJul 27th 2014 edited
     
    Ok guys, I just finished building my first computer and I did the benchmark with Vekoma Fan Boy's Settings with the exception of the monitor being 1680x1050
    nolimits2app 2014-07-26 18-53-55-02
    CPU: AMD Athlon x4 760k 3.8GHz
    RAM: 4 GBS Kingston HyperX Fury 1600MHz
    GPU: XFX Radeon HD 5870
    • CommentAuthorJetPulse
    • CommentTimeJul 27th 2014
     
    Laptop: 17.03 fps (generally what I'm using when working on NL projects)
    Intel i7 8 core @ 1.65GHz
    6GB RAM
    Nvidia GTM230

    Desktop: 72.60 fps (hardly run NL on it but mainly use it for terrain editing scripts or to do heavy lifting 3ds max functions)
    Intel i5 4 core @ 3.4GHz
    8GB Ram
    Nvidia GTX660
  13.  
    i tried to benchmark but when i went into the properties there is no shortcut tab
    •  
      CommentAuthorUncleArly
    • CommentTimeJul 30th 2014
     
    ^It's the default tab when you open properties on any Windows OS.
    • CommentAuthorROLLER97
    • CommentTimeJul 30th 2014
     


    Intel Core i7 2600K @4.39 GHz
    16 GB RAM
    MSI GeForce GTX 660 TF OC
    1280x1024 resolution
    • CommentAuthorJetPulse
    • CommentTimeJul 30th 2014
     
    Make sure you are actually clicking on a shortcut and not on the NL2 executable. If you dig into your NL2 folder you will find the actual file and that will not have a shortcut menu. (Assuming you're on Windows) If you don't have a shortcut, you can right click the executable, go to "send to" and click "Desktop (create shortcut)" and it will create the proper file on your desktop that you can change as shown in the tutorial above. You should also be able to use the above method on the NL2 listed in your start menu (start menu Icons behave exactly like shortcuts).
  14.  
    oh, that was my problem, thanks

    average fps: 18.33

    Intel core i5 4210U @ 1.7 Ghz
    6 GB RAM DDR3 1600 mhz
    Nvidia GT 820M 2 GB
    1920x1080 screen
  15.  
    Yeah - I have two NL2 shortcuts on my desktop now, one for the 'real' game and one for the benchmark version.
  16.  
    Just redid it:

    RESOLUTION: 1920x1080
    Avg FPS: 88.13
    RESOLUTION: 4400x1080
    Avg FPS: 59.78

    AMD 6300 SIX-CORE PROCESSOR
    8GB RAM DDR3
    NVIDIA (EVGA) GEFORCE GTX 770 3GB
    • CommentAuthorDistortAMG
    • CommentTimeJul 31st 2014 edited
     
    I have 2 different screen shots to show you, as I have two graphics adapters inside my laptop. One is integrated into my AMD APU and the other is dedicated. I had to run the benchmark at the laptops native resolution, trying to force a higher resolution through the catalyst control panel presented me with a blank screen. The standardization for the benchmarks that someone suggested in an earlier post have also been used for my results.

    My specs:
    HP Pavillion N268SA Laptop
    AMD A10-5745M Quad Core 2.1 - 2.9 GHZ APU
    8GB DDR3 1600MHZ
    AMD Radeon HD 8610G Integrated into APU (768MB Shared) 533 - 626MHZ Clock, 384 Shader processors, 128 bit memory bus width.
    AMD Radeon HD 8670M Dedicated (2GB DDR3 Dedicated) 775 - 900MHZ Clock, 384 Shader processors, 64 bit memory bus width.
    Windows 8.1 with update 1 (64bit)
    1366 x 768 Native resolution
    AMD Catalyst 14.7 Release candidate drivers
    High performance power mode set


    Here is the first screen shot, running on the AMD Radeon HD 8610G. 1366 x 768
     photo 8550G_zpsf3efb301.png

    Here is the second screen shot, running on the AMD Radeon HD 8670M. 1366 x 768
     photo 8670m_zps0ea39400.png

    Interestingly, the integrated graphics chip performs better than the dedicated graphics chip, despite the dedicated graphics chip running at higher clock speeds. Both chips have 384 shader processors. I suspect that this lies with the memory bandwidth of the chips, the 8610G has 128bit memory bus bandwidth, where the dedicated 8670M only has a 64 bit bus bandwidth. Giving peak throughput to the memory of 21.3GB/s and 14.4GB/s respectively. The texture fill rate is also approximately a third higher with the integrated graphics than what it is with the dedicated graphics, I would have thought that this is a direct result of the memory bandwidth.
    Seeing as this system architecture does support dual graphics, an application profile for NL2 would be nice so it runs in dual graphics mode. I would have thought this would need to be created by AMD though as it comes in their drivers. So cannot see that happening. Mantle API support would be nice too seeing as my 8670M architecture supports it, but I think my computer while running NL2 is more GPU limited and not CPU limited, which would make it useless. I am also sure that the time spent developing and supporting those things could be much better spent else where, like making this fantastic simulator even more awesome :)

    I also find Mr. Laptop's results quite interesting, as he is running a AMD A6-4400M APU, which is a dual core, but is not classed as a true dual core CPU due to some of the major core components in the CPU being shared. Now my CPU, is a quad core but it is not a true quad core. Because, (like the Intel Core 2 Quad's) It is two dual core CPU's on the same die with some of the supporting components for the main core components being shared. Also, his integrated GPU has a lower clock speed, and less shader processors, 192 versus my integrated GPU's 384. But, he does have a slightly better memory bandwidth. Yet if his results are correct he gets approx 12.4FPS. Which is just slightly lower than my integrated benchmark. Obviously, there are other factors at play, but very interesting indeed.

    Finally, running the simulator with most settings on medium, and textures on maximum gives me a very playable and usable 45FPS on average on the same park. Very happy with that result. It will be interesting to see what my results will be when and if I purchase the monster of a computer I have been looking at, hopefully this side of Christmas. But with my son just being born, money and time is very tight at the moment.
    •  
      CommentAuthoraw555000
    • CommentTimeJul 31st 2014
     
    ^^ Your specs are really similar to mine, I'm just A8 not A10 and I'm windows 7 not 8.
    • CommentAuthorDistortAMG
    • CommentTimeJul 31st 2014 edited
     
    It would be interesting to see what your benchmarks are, especially if your A8 is one of the mobile variants.

    I do have another laptop with an AMD A8 5557M APU inside, paired with an AMD Radeon HD 8550G integrated graphics. This has 256 shader processors inside.

    My AMD A8 and A10 are based off the newer Richland architecture and Mr. Laptops A6 is based off the Trinity architecture I believe, there is not many changes between the two besides power consumption and slight clock speed changes. All 3 chips mentioned above are also based off the reworked Bulldozer microarchitecture called Piledriver. So it could be an even match for a fairly fair comparison between a A6, A8, and A10 mobile APU, with 192, 256, 384 shader processors in their integrated GPU's respectively.

    I do not want to go installing NL2 on a second computer and risk getting my licence revoked however, I am not sure what Ole's policy on this is though, and if he allows it on a second computer. So unless I am told specially by a person with the authority to tell me, that it is allowed. I wont be temporarily installing it on that system just for the purpose of a benchmark. If it is allowed however, I will try get a benchmark up within the next day or two.

    As I said, with my A10, on medium graphic settings I get on average approximately 45FPS on Forest Mountain Park. While it is no match for some of the beasts on this page. The laptop cost less than £500, and is portable, win win situation really.
    •  
      CommentAuthoraw555000
    • CommentTimeJul 31st 2014
     
    I will run the benchmark test tomorrow and post my results, it would be nice to see a direct comparison between the A6, A8 and A10.
    Yeah mines mobile to be exact its AMD Quad-Core A8-3520M (1.6 GHz, 4 MB L2 cache) with AMD Radeon HD 6620G + 7450M Dual GPU and 6GB of RAM.
    You seem like someone who knows there tech...respect
  17.  
    Think it's time for an upgrade...
    fps

    With settings down to a minimum the results were somewhere around 18 fps and 1100 frames

    Specs:
    HP Omni 100 PC
    AMD Athlon II X2 260u Processor 1.80 GHz
    4GB RAM
    •  
      CommentAuthoraw555000
    • CommentTimeJul 31st 2014 edited
     

    I'm not that disappointed, NL runs pretty well for me most of the time sticking between 30-50FPS on most coasters with default settings.

    Laptop Specs:
    Model: HP Pavilion G6-1331SA
    CPU: AMD Quad-Core A8-3520M (1.6 GHz, 4MB L2 cache)
    GPU: AMD Radeon HD 6620G + 7450M Dual GPU
    RAM: 6GB
    Resolution: 1366 x768
    OS: Windows 7 SP1
    Good laptop considering it cost £350 almost 2 years ago and plays everything I want it to really well (including the Sims 3, The Movies and obviously NL).

    @DistortAMG I find it weird that I've got a slightly higher FPS than your A10 with my A8. What settings did you use? I'll redo with your settings.
    • CommentAuthorole
    • CommentTimeJul 31st 2014 edited
     
    Installing NL2 on multiple computers is allowed, as long as all the computers belong to the same person and they are operated by that person only. If you install it on multiple computers and those computers are operated by multiple people, you would require a separate license for each person.

    Number of licenses that are required = Minimum of (Number of computers and Number of persons who you use the computers)
    • CommentAuthorDistortAMG
    • CommentTimeJul 31st 2014 edited
     
    Thank you for that clarification Ole, I installed it onto my second computer for the purpose of benchmarks and have now removed it. As I only use the second laptop for light activities and programming.

    Here are the specs of my second laptop:

    Acer Aspire V5-552
    AMD A8-5557M Quad Core 2.1 - 3.1 GHZ APU
    8GB DDR3 1600MHZ
    AMD Radeon HD 8550G Intergrated into APU (768MB Shared memory) 554 - 720 MHZ Clock, 256 Shader processors, 128bit memory bus width
    Windows 8.1 with update 1 (64bit)
    1366 x 768 Native resolution
    AMD Catalyst 14.7 Release candidate drivers
    High performance power mode set


    Here are the results
     photo Acer_zps8564f5ac.png


    @aw555000, Interesting indeed, I believe your integrated chip is the most powerful of your two, and which I assume is what is being used. Having 400 shader processors and 128bit memory bus, this could by why your results are higher than mine. Although your CPU is slower, this could be more evidence to indicate that on our systems we are GPU rather than CPU limited.

    My benchmark settings for all 3 benchmarks were:

    High power mode set in Windows power options.
    1366 x 768 Resolution in NL2 (Windowed mode)
    Triple Buffering on
    Anti Aliasing set to Level 3

    Every slider in the detail tab set to maximum
    Advanced effects NOT selected