Skip to content
TheChatArchiveThe Chat Archive
gpt-4-1106-preview· OpenAIAi2 WildChat-4.8M (AI2)

Can you give me a dxvk.conf for games that are bottlenecked by CPU like world of…

  1. User

    Can you give me a dxvk.conf for games that are bottlenecked by CPU like world of warcraft 3.3.5a private server that uses directx9 32 bit Here are my specs: CPU: AMD Ryzen 9 7950X 4.5 GHz 16-Core Processor 32-Thread Motherboard: MSI MPG B650I EDGE WIFI Mini ITX AM5 Motherboard Memory: Mushkin Enhanced Redline 64 GB (2 x 32 GB) DDR5-5600 CL28 Memory Storage: Samsung 990 Pro w/Heatsink 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive Storage: Western Digital Black SN850X 4 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive Video Card: Sapphire PULSE Radeon RX 7900 XT 20 GB Video Card Power Supply: Corsair SF1000L 1000 W 80+ Gold Certified Fully Modular SFX Power Supply Rear Fan: Noctua A9 PWM chromax.black.swap 46.44 CFM 92 mm Fan Radiator Fan: be quiet! Silent Wings Pro 4 97.41 CFM 140 mm Fan Keyboard: Keychron K3 RGB Wired Gaming Keyboard Mouse: SteelSeries Aerox 9 Wireless Optical Mouse @ 3100 dpi Headphones: Sony XM5 Headset Microphone: HyperX QuadCast S – RGB USB Condenser Microphone DAC: FiiO New K3 Headphone Amplifier 384kHz/32bit Monitor: LG OLED Evo C2 Series 42” 4K Smart TV (3840 x 2160), 120Hz Refresh Rate AMD freesync 800 nits with HDR OS: Windows 11 23H2

    here is the dxvk i want you to change:

    Supported values: True, False

    dxgi.enableHDR = True

    Supported values: True, False

    dxgi.deferSurfaceCreation = False

    d3d9.deferSurfaceCreation = False

    Supported values : 0 - 16

    dxgi.maxFrameLatency = 0

    d3d9.maxFrameLatency = 0

    Supported values : Any non-negative integer

    dxgi.maxFrameRate = 0

    d3d9.maxFrameRate = 0

    Supported values: Any four-digit hex number.

    dxgi.customDeviceId = 0000

    dxgi.customVendorId = 0000

    d3d9.customDeviceId = 0000

    d3d9.customVendorId = 0000

    Supported values: Any string.

    dxgi.customDeviceDesc = “”

    d3d9.customDeviceDesc = “”

    Supported values: Auto, True, False

    dxgi.hideNvidiaGpu = Auto

    Supported values: Auto, True, False

    dxgi.hideAmdGpu = Auto

    Supported values: Auto, True, False

    dxgi.hideIntelGpu = Auto

    Supported values: Any number in Megabytes.

    dxgi.maxDeviceMemory = 0

    dxgi.maxSharedMemory = 0

    Supported values: True, False

    dxgi.emulateUMA = False

    Supported values: Any number greater than or equal to 2.

    dxgi.numBackBuffers = 0

    d3d9.numBackBuffers = 0

    Supported values: Any non-negative number

    dxgi.syncInterval = -1

    d3d9.presentInterval = -1

    Supported values: Auto, True, False

    dxvk.tearFree = Auto

    Supported values: Any number between 0 and 16

    d3d9.samplerAnisotropy = -1

    Supported values: Any number between -2.0 and 1.0

    d3d9.samplerLodBias = 0.0

    Supported values: True, False

    d3d9.clampNegativeLodBias = False

    Supported values: True, False

    d3d9.invariantPosition = True

    Supported values: True, False

    d3d9.forceSampleRateShading = False

    Supported values:

    - 0 to use all available CPU cores

    - any positive number to enforce the thread count

    dxvk.numCompilerThreads = 0

    Supported values:

    - Auto: Don’t change the default

    - True, False: Always enable / disable

    dxvk.useRawSsbo = Auto

    Supported values:

    - 0 to use the defaults

    - any positive integer to limit the chunk size, in MiB

    dxvk.maxChunkSize = 0

    Supported values:

    - Auto: Enable if supported, and compile optimized pipelines in the background

    - True: Enable if supported, but do not compile optimized pipelines

    - False: Always disable the feature

    dxvk.enableGraphicsPipelineLibrary = Auto

    Supported values:

    - Auto: Enable tracking for 32-bit applications only

    - True: Always enable tracking

    - False: Always disable tracking

    dxvk.trackPipelineLifetime = Auto

    Supported values:

    - 1: Shader Model 1

    - 2: Shader Model 2

    - 3: Shader Model 3

    d3d9.shaderModel = 3

    Supported values:

    - True, False: Always enable / disable

    d3d9.dpiAware = True

    Supported values:

    - True, False: Always enable / disable

    d3d9.strictConstantCopies = False

    Supported values:

    - True, False: Always enable / disable

    d3d9.strictPow = True

    Supported values:

    - True, False: Always enable / disable

    d3d9.lenientClear = False

    Supported values:

    - Max Available Memory: Any int32_t

    - Memory Tracking Testing: True, False

    d3d9.maxAvailableMemory = 4096

    d3d9.memoryTrackTest = False

    Supported values:

    - True: Use a faster but less accurate approach. Good enough for most games

    - False: Disable float emulation completely

    - Strict: Use a slower but more correct approach. Necessary for some games

    - Auto: DXVK will pick automatically

    d3d9.floatEmulation = Auto

    Supported values:

    - True, False: Always enable / disable

    d3d9.enableDialogMode = False

    Supported values: -1 (application) and 0 to 16 (user override)

    d3d9.forceSwapchainMSAA = -1

    Supported values:

    - True/False

    d3d9.longMad = False

    Supported values:

    - True/False

    d3d9.deviceLocalConstantBuffers = False

    Supported values:

    - True/False

    d3d9.supportDFFormats = True

    Supported values:

    - True/False

    d3d9.useD32forD24 = False

    Supported values:

    - True/False

    d3d9.supportX4R4G4B4 = True

    Supported values:

    - True/False

    d3d9.supportD32 = True

    Supported values:

    - True/False

    d3d9.disableA8RT = False

    Supported values:

    - True/False

    Defaults to True if vendorId == 0x10de

    d3d9.supportVCache = True

    Supported values:

    - True/False

    d3d9.forceSamplerTypeSpecConstants = False

    Supported values:

    - Any ratio, ie. “16:9”, “4:3”

    d3d9.forceAspectRatio = “”

    Supported values:

    - True/False

    d3d9.enumerateByDisplays = True

    Supported values:

    - True/False

    d3d9.cachedDynamicBuffers = False

    Supported values:

    - True/False

    d3d9.seamlessCubes = False

    Supported values:

    - True/False

    dxvk.enableDebugUtils = False

    Supported values:

    - True/False

    dxgi.useMonitorFallback = False

    Supported values:

    - True/False

    dxvk.hideIntegratedGraphics = False

    please refrain from commenting

  2. Assistant

    The configuration provided is comprehensive, and many of the settings you've included are disabled or set to follow default behavior. You can often leave these settings as default for most games, especially considering your high-end system specs.

    World of Warcraft (WoW) 3.3.5a is a relatively old version of the game, and even though you're playing on a private server, generally your CPU should not be a bottleneck, especially a Ryzen 9 7950X.

    For WoW on a private server, less tweaking might be needed, but since DXVK converts Direct3D 9 to Vulkan for better performance on Linux, if you run into any specific performance issues like stuttering or CPU bottlenecks, you can try making a few changes. Here's a modified version focusing on potential performance improvements for your DXVK configuration:

    d3d9.numCompilerThreads = 16  # Use 16 threads for shader compilation to spread the load
    d3d9.maxFrameLatency = 1      # Reduce input lag by limiting frame latency
    
    # Can help with stability and stuttering in some games
    dxvk.enableGraphicsPipelineLibrary = False
    
    # Use all available CPU cores efficiently, adapt number to your CPU thread count if necessary
    dxvk.numCompilerThreads = 0
    
    # May improve performance on systems where memory is not an issue, set to VRAM size in MB
    dxgi.maxDeviceMemory = 20480   # Assuming you have 20GB GPU VRAM
    dxgi.maxSharedMemory = 32768   # Assuming you're comfortable with up to 32GB shared system memory usage
    
    # Shader-related tweaks typically aren't necessary on high-end hardware, but these can sometimes offer improvements
    d3d9.shaderModel = 3          # Should be adequate for WoW 3.3.5a
    d3d9.strictPow = False         # More performance-oriented setting for power operations in shaders
    d3d9.floatEmulation = False    # Since your hardware is modern, there should be no need for float emulation
    

    Most other settings can remain at their defaults, as you're running a high-end system and these configurations might have little to no noticeable effect, or could even introduce problems if not tailored to address specific issues you encounter. If you're looking for very specific tweaks related to WoW 3.3.5a, the game's community or forums might have dedicated threads detailing the best settings for performance.

    Remember, DXVK is typically used on Linux. If you're on Windows (as your included specs suggest Windows 11), DXVK might not be the best solution, unless you've specifically set up your system to use it.

    Ensure that your WoW 3.3.5a private server game client is compatible with DXVK and/or D9VK. Normally, CPU bottleneck issues are not expected with your hardware, so make sure in-game settings (like view distance, details, etc.) are set appropriately. If you are indeed playing through WINE or a similar compatibility layer on Linux, make sure you have the latest versions of these tools for optimal performance.

  3. User

    Is this good?

    d3d9.maxFrameRate = 0 d3d9.maxFrameLatency = 1 d3d9.numBackBuffers = 0 d3d9.presentInterval = 0 d3d9.tearFree = False d3d9.maxAvailableMemory = 16384 d3d9.evictManagedOnUnlock = True d3d9.allowDiscard = True dxvk.enableAsync = True dxvk.numCompilerThreads = 0 d3d9.samplerAnisotropy = 16 d3d9.invariantPosition = False d3d9.memoryTrackTest = False d3d9.noExplicitFrontBuffer = False d3d9.strictConstantCopies = False d3d9.lenientClear = True dxvk.maxChunkSize = 256 d3d9.longMad = False d3d9.floatEmulation = False
    d3d9.forceSwapchainMSAA = 0 d3d9.supportVCache = True d3d9.forceSamplerTypeSpecConstants = False dxvk.useRawSsbo = False d3d9.disableA8RT = True d3d9.shaderModel = 3 d3d9.dpiAware = True numCompilerThreads = 0 dxgi.maxDeviceMemory = 20480 d3d9.strictPow = False

  4. Assistant

    Your modified dxvk.conf configuration generally looks good for a high-end system such as yours, aiming to achieve high frame rates without unnecessary limitations. However, there are a few points to keep in mind or potentially adjust according to your needs:

    1. d3d9.maxFrameRate = 0 - This sets no cap on the frame rate. Unless you're experiencing issues with too high frame rates (e.g., coil whine, overheating, etc.), this is fine.

    2. d3d9.maxFrameLatency = 1 - Lower latency helps with responsiveness but can increase CPU load because it may queue fewer frames.

    3. d3d9.presentInterval = 0 - Disabling VSync can reduce input lag but might cause screen tearing. Since your monitor supports Freesync, you might want to take advantage of that instead.

    4. dxvk.enableAsync = True - Be cautious with this option, as it is not officially supported by DXVK, and can cause instability, crashes, or graphical issues. It's often used to reduce stuttering due to shader compilation.

    5. d3d9.samplerAnisotropy = 16 - Forces max anisotropic filtering, which can improve texture quality at oblique angles.

    6. dxvk.numCompilerThreads = 0 - This setting makes DXVK use all logical CPU cores for shader compilation. Considering your powerful CPU, this is usually fine. However, duplicate this line as numCompilerThreads = 0 seems to be unintended.

    7. dxvk.maxChunkSize = 256 - This setting limits the chunk size for state cache files. A higher number might improve game loading times but can consume more memory.

    8. d3d9.floatEmulation = False - Turning off float emulation can improve performance but might cause issues in some games that rely on D3D9's float behavior.

    9. dxgi.maxDeviceMemory = 20480 - This sets the maximum amount of dedicated GPU memory DXVK will report to games. It's set correctly for your GPU.

    10. d3d9.strictPow = False - A performance-oriented setting, but can result in different visual outcomes for the pow HLSL function.

    11. The option dxvk.useRawSsbo seems to be omitted from your configuration. It's worth including it if you have issues with games that use SSBOs.

    12. Similarly, options related to HDR (dxgi.enableHDR) are missing, which you might want to enable if you're using an HDR-capable monitor.

    Remember, configs like this should be tailored to solve specific issues. If you're not encountering any problems, the default settings may already be optimal. Always test your configuration in-game to ensure it's functioning as intended, and be ready to revert back to defaults or to adjust specific options if you encounter any unwanted behavior.

    Moreover, please ensure you are actually using DXVK in your system. The configuration you have won't have any effect if you're running games in Windows without a wrapper that uses DXVK to translate D3D9 calls to Vulkan. DXVK is predominantly used in a Linux environment.

    Lastly, some private servers might have specific rules about client modifications, it's wise to ensure any changes you make to the game's operation comply with their rules to avoid any potential issues.

  5. User

    Should i change this??

    d3d9.strictConstantCopies = False

  6. Assistant

    The d3d9.strictConstantCopies option controls whether DXVK should perform strict data validation when updating constant buffers. This can be particularly relevant when running older DirectX 9 games that might mismanage constant buffer updates, causing rendering issues.

    Setting d3d9.strictConstantCopies to False may result in a slight performance gain since it reduces some of the extra validation overhead. This could potentially help in CPU-bound scenarios. However, it could also introduce graphical artifacts if the game relies on the exact D3D9 behavior that this validation checks for.

    You have a very capable CPU (AMD Ryzen 9 7950X), so the performance gain from setting this option to False might be negligible. In this case, the safe choice would be to leave this setting as the default (True), which ensures that constant buffer updates follow the strict rules the game might be expecting based on the original D3D9 behavior.

    In summary, you should generally keep d3d9.strictConstantCopies set to True unless you actually encounter performance issues and suspect this setting might help. As always, after adjusting these settings, it is advisable to test the game to make sure there are no adverse effects on graphics or stability. If you do see improvements with no negative side effects, then keeping it as False can be the better choice for your particular use case.

  7. User

    What would be optimal for this game?

    Changes memory chunk size.

    Can be used to override the maximum memory chunk size.

    Supported values:

    - 0 to use the defaults

    - any positive integer to limit the chunk size, in MiB

    dxvk.maxChunkSize = 0

  8. Assistant

    The dxvk.maxChunkSize configuration option in DXVK specifies the maximum size of memory chunks for resources like textures and buffers. When DXVK has to allocate memory, it does so in chunks, and this setting allows you to control the size of those chunks.

    The default value of 0 lets DXVK choose the chunk size automatically based on what it considers optimal. Typically, the default setting is fine for most users, and it should work well, particularly as DXVK's defaults are made to suit a wide range of hardware and game requirements.

    Given your high-end specifications (particularly the large amount of VRAM on your Sapphire PULSE Radeon RX 7900 XT 20 GB Video Card), you shouldn't need to adjust this setting unless you're running into specific issues that suggest memory allocation problems, such as stuttering or crashes related to memory.

    However, if you experience such issues or want to experiment with performance tweaking, you could set the dxvk.maxChunkSize to a value in megabytes. Some users report performance improvements by setting it to values between 64 and 256.

    Remember to monitor the game's performance while making these adjustments. If you notice no difference or experience a decrease in performance or stability, it's best to revert to the default (0). As is often the case with configuration settings like this, there is no universal "optimal" number—it can vary from game to game or system to system.

    In summary, for a game like World of Warcraft 3.3.5a running on a robust system, you are likely safe leaving this setting at its default. Only consider changing it if you're troubleshooting a specific issue that could be related to memory chunk allocation.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.