Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Topics - JeGX

Pages: [1] 2 3 ... 55
Geeks3D's GPU Tools / GPU Caps Viewer 1.42.0 released
« on: March 22, 2019, 06:50:16 PM »
GPU Caps Viewer has been released.

Release notes and download links:

Version - 2019.03.22
+ added alpha version of GPU temperature and core usage reading
  on Radeon VII (Vega20 / Vega12 GPUs) - Not tested.
+ added support of AMD Radeon VII.
+ added support of NVIDIA GeForce GTX 1660 Ti and
  GeForce GTX 1660.
+ added support of Intel UHD 630 (0x3e98).
! updated Vulkan Tessellation v2 demo (improved GPU utilization and
  recompilation of GLSL shaders to SPIRV).
* fixed a bug in the pixel shader of the Vulkan Seascape
  demo that caused visual artifacts on Intel GPUs.
* fixed saving of bench_result.csv used for for OpenGL and
  OpenCL demos (see HERE).
! updated with latest GeeXLab SDK libs.
! updated: GPU Shark
! updated: ZoomGPU 1.23.0 (GPU monitoring library).

GPU Caps Viewer + GeForce GTX 1080

Geeks3D's GPU Tools / GPU Sharl released
« on: March 22, 2019, 06:42:57 PM »
A new update of GPU Shark is available.

Download and release notes:

GPU Shark

3D-Tech News Around The Web / NVIDIA PhysX SDK 4.1 released
« on: March 21, 2019, 04:10:30 PM »
Release Notes - NVIDIA® PhysX® SDK 4.1.0


    Immediate mode support for reduced-coordinates articulations and the temporal Gauss Seidel solver.
    GPU acceleration for reduced-coordinates articulations.


        Added support for UWP, note that cmake 3.13.4 is required for uwp arm64.
        PhysXGpu DLLs are now standalone, so they will now work with both static and dynamic PhysX libraries.
        PhysX delay loading code is disabled for the static library configuration.
        Removed PxGpuDispatcher class. Instead of querying the GPU dispatcher with PxCudaContextManager::getGpuDispatcher() and providing it to the PxScene with PxSceneDesc::gpuDispatcher, please provide the CUDA context manager directly using PxSceneDesc::cudaContextManager.
        PxCreateCudaContextManager does have an additional parameter PxProfilerCallback, that is required in order to get profiling events from the GPU dll.
        FastXml project is now compiled as OBJECT on win platforms and is linked into PhysXExtensions library.
        Removed PxArticulationBase::getType(), PxArticulationBase::eReducedCoordinate, PxArticulationBase::eMaximumCoordinate and added PxConcreteType::eARTICULATION_REDUCED_COORDINATE, PxConcreteType::eARTICULATION_JOINT_REDUCED_COORDINATE.

Rigid Bodies

        Immediate mode API for reduced-coordinates articulations.
        Immediate mode API for the temporal Gauss Seidel (TGS) solver .
        Compute dense Jacobian matrix for the reduced-coordinates articulations.
        GPU acceleration for reduced-coordinates articulations with PGS solver.
        GPU acceleration for reduced-coordinates articulations with TGS solver (experimental).
        PxSimulationStatistics::nbDynamicBodies does not include kinematics any longer. Instead they are covered in new nbKinematicBodies counter.
        Fixed speculative CCD optimization with sleeping bodies.
        Fixed the overlap termination condition in the GJK code for sphere primitive.
        Fixed a bug in the face selection algorithm for paper thin box overlapped with capsule.
        Fixed a contact recycling issue with PCM contact gen.
        Fixed an issue with articulation when removing and adding links to the articulation.


        PxSerialization::serializeCollectionToBinaryDeterministic, convenience function to post-process binary output with PxBinaryConverter for determinism. For achieving determinism, the checked build needs to be used.
        Support for binary and xml serialization for PxArticulationReducedCoordinate.
        PxBinaryConverter can now produce deterministic output, independent of the runtime environment the objects have been serialized in. For achieving determinism, the checked build needs to be used for serializing collections.
        PX_BINARY_SERIAL_VERSION has been changed to a global unique identifier string. PX_PHYSICS_VERSION is no longer part of binary data versioning.


3D-Tech News Around The Web / NVIDIA - Ray Tracing Best Practices
« on: March 21, 2019, 03:20:14 PM »
This post presents best practices for implementing ray tracing in games and other real-time graphics applications. We present these as briefly as possible to help you quickly find key ideas. This is based on a presentation made at the 2019 GDC by NVIDIA engineers.


Real Time Ray Tracing

NVIDIA has released a new family of graphics drivers called Creator Ready Drivers (CRD).

NVIDIA RTX GPUs have given video editors, 3D animators, broadcasters, graphic designers and photographers around the world a major boost in productivity with hardware-accelerated ray tracing, AI-enhanced creative workflows and advanced video processing.

But as demands increase on creators to generate higher volume of quality content, it’s more important than ever to ensure the apps and hardware that creators depend on never miss a beat. So NVIDIA has introduced a Creator Ready Driver program to provide artists, creators and game developers the best performance and reliability when working with creative applications.

To achieve the highest level of reliability, Creator Ready Drivers undergo extensive testing against multiple revisions of the top creative applications. Whether you just updated to the latest release of your favorite creative app or are still on the prior version, Creator Ready Drivers are fully tested.

Complete story:

NVIDIA Creator Ready Drivers

- latest CRD win10 64-bit @ Geeks3D
- 419.67 win10 64-bit @ NVIDIA

Supported Graphics Cards


GeForce RTX 20 Series:
- GeForce RTX 2080 Ti, GeForce RTX 2080, GeForce RTX 2070, GeForce RTX 2060

GeForce 16 Series:
- GeForce GTX 1660 Ti, GeForce GTX 1660

GeForce 10 Series:
- GeForce GTX 1080 Ti, GeForce GTX 1080, GeForce GTX 1070 Ti, GeForce GTX 1070, GeForce GTX 1060, GeForce GTX 1050 Ti, GeForce GTX 1050

CRD 419.67 exposes  OpenGL 4.6 (419 extensions) and Vulkan 1.1.95 for a GeForce RTX 2080.

GPU Caps Viewer + GeForce RTX 2080

The C++ compiler in Visual Studio 2019 includes several new optimizations and improvements geared towards increasing the performance of games and making game developers more productive by reducing the compilation time of large projects. Although the focus of this blog post is on the game industry, these improvements apply to most C++ applications and C++ developers.


DXR support on GTX 10 / 16 Series not before April 2019.

When Microsoft first announced the DirectX Raytracing API just over a year ago, they set out a plan for essentially two tiers of hardware. The forward-looking plan (and long-term goal) was to get manufacturers to implement hardware features to accelerate raytracing. However in the shorter term, and owing to the fact that the API doesn't say how ray tracing should be implemented, DXR would also allow for software (compute shader) based solutions for GPUs that lack complete hardware acceleration. In fact, Microsoft had been using their own internally-developed fallback layer to allow them to develop the API and test against it internally, but past that they left it up to hardware vendors if they wanted to develop their own mechanisms for supporting DXR on pre-raytracing GPUs.

NVIDIA for their part has decided to go ahead with this, announcing that they will support DXR on many of their GeForce 10 (Pascal) and GeForce 16 (Turing GTX) series video cards. Specifically, the GeForce GTX 1060 6GB and higher, as well as the new GTX 1660 series video cards.

Now, as you might expect, raytracing performance on these cards is going to be much (much) slower than it is on NVIDIA's RTX series cards, all of which have hardware raytracing support via NVIDIA's RTX backend. NVIDIA's official numbers are that the RTX cards are 2-3x faster than the GTX cards, however this is going to be workload-dependent. Ultimately it's the game and the settings used that will determine just how much of an additional workload raytracing will place on a GTX card.

Complete story:

The Radeon GPU Analyzer (RGA) is an offline compiler and code analysis tool for DirectX shaders, OpenGL shaders, Vulkan shaders and OpenCL kernels.

Radeon GPU Analyzer is a compiler and code analysis tool for Vulkan, DirectX, OpenGL and OpenCL. Using this product, you can compile source code for a variety of AMD GPUs and APUs, independent from the GPU/APU that is physically installed on your system, and generate GCN ISA disassembly, intermediate language disassembly, performance statistics and static analysis reports for each target platform.

This product can be used to produce the following output:
- AMD GPU ISA disassembly for different GCN generations
- Intermediate langauge disassebly: AMDIL, HSAIL, DX ASM for DirectX shaders, LLVM IR for ROCm OpenCL, SPIR-V for Vulkan
- Compiled binaries
- Performance statistics
- Live register analysis (see for more info)
- Control-Flow graphs
- Build errors and warnings

The RGA package contains both a GUI app and a command-line exectuable.

The supported platforms by the GUI app are:
- Vulkan - GLSL/SPIR-V as input, together with the Vulkan pipeline state, through AMD's Vulkan driver
- OpenCL - AMD's LLVM-based Lightning Compiler for ROCm

The supported platforms by the command-line tool are:
- D3D11
- OpenCL - AMD's LLVM-based Lightning Compiler for ROCm (-s rocm-cl)
- OpenCL - Legacy compiler (-s cl)
- OpenGL
- Vulkan - Compilation of GLSL/SPIR-V together with the API's pipeline state, using AMD's Vulkan driver
- Offline Vulkan - Using a static compiler, accepts GLSL/SPIR-V as input. Note that the this mode is about to be removed from RGA. To ensure that the results that RGA provides are accurate and reflect the real-life case, please use the new Vulkan live driver mode (which is also supported in the GUI application).


Radeon GPU Analyzer

libfacedetection is an open source library for face detection in images. The face detection speed can reach 1500FPS.

This is an open source library for CNN-based face detection in images. The CNN model has been converted to static variales in C source files. The source code does not depend on any other libraries. What you need is just a C++ compiler. You can compile the source code under Windows, Linux, ARM and any platform with a C++ compiler.

SIMD instructions are used to speedup the detection. You can enable AVX2 if you use Intel CPU or NEON for ARM.

The model file has also been provided in directory ./models/.

examples/libfacedetectcnn-example.cpp shows how to use the library.


libfacedetection demo

3D-Tech News Around The Web / GPU-Z 2.18.0 released
« on: March 15, 2019, 09:41:53 AM »
- Download GPU-Z @ Geeks3D
- Download GPU-Z @ TPU


Changelog v2.18.0
    Added support for NVIDIA GeForce GTX 1660, RTX 2060 Mobile (TU106-B), Quadro RTX 5000, Quadro RTX 6000
    Added Hotspot temperature monitoring for AMD Radeon VII
    Vega 10 & Vega 20 Hotspot no longer disabled by default
    Updated PCI vendor database to add new companies and remove unused entries
    Ensure "Sapphire" vendor name doesn't get displayed on PCPartner OEM cards
    Fixed crash on startup during BIOS parsing on older NVIDIA cards (GeForce 500)
    Fixed GPU-Z hang during splash screen on some systems
    Fixed system crash on some AMD systems during BIOS read

GPU-Z 2.18.0 + GeForce GTX 1080

The Radeon RX 560 XT is a cut down version of the RX 570 and fills the gap between the RX 560 and the RX 570.

RX 560 XT main features:
- GPU Polaris 10
- 1792 shader cores
- 28 computes units
- 112 TMUs
- 32 ROPs
- 4GB or 8 GB of GDDR5 memory @ 6.6 Gbps, 256-bit
- 150 W TDP

Sapphire is the unique vendor on the Chinese market.

More info:

Radeon RX 560 XT

What’s New in Version 419.35 WHQL 
- Provides the optimal gaming experience for Apex Legends, Devil May Cry 5, and Tom Clancy's The Division II.

- Includes support for three new G-SYNC compatible monitors (monitors that deliver a baseline Variable Refresh Rate (VRR) experience):
  - Acer ED273 A
  - Acer XF250Q
  - BenQ XL2540-B / ZOWIE XL LCD

Fixed Issues in this Release
- [G-SYNC]: With a G-SYNC and G-SYNC Compatible display connected in clone mode, flashing occurs on games played on the G-SYNC display with G-SYNC enabled. [200482157]
- [Apex Legends]: DXGI_ERROR_DEVICE_HUNG error message may appear and then the game crashes.[2503360]
- [Hitman 2]: Pixelated corruption appears in the game. [2504274]
- [Batman: Arkham Origins]: PhysX fog renders incorrectly [2451459]
- [GeForce RTX 2080][PhysX][Assassin’s Creed 4 - Black Flag]: Smoke dispersal appears accelerated. [2498928]
- Microsoft.Photos.exe randomly crashes. [200496899]
- NVDisplay.Container.exe causes high CPU usage. [200496099]

- 419.35 @ Geeks3D
- 419.35 win10 (desktop) @ NVIDIA
- 419.35 win7/win8 (desktop) @ NVIDIA

GeForce 419.35 exposes OpenGL 4.6 (419 extensions) and Vulkan 1.1.95 like previous 418.81.

More information about this new driver can be found here:
- Our Latest Game Ready Driver Optimizes Apex Legends, Devil May Cry 5 and The Division II, and Adds Support For New G-SYNC Compatible Gaming Monitors

GPU Caps Viewer + GeForce RTX 2080

3D-Tech News Around The Web / Vulkan SDK 1.1.101 released
« on: March 04, 2019, 06:11:10 PM »
New Vulkan Extensions plus new look for Vulkan SDK Download Site
The Vulkan SDK Download site has a new look with Vulkan branding. Check out the site and download the new Vulkan SDKs for Windows, Linux, and MacOS that include the most recent extensions. More info about the new SDK follows.

LunarG releases new Vulkan SDK for header
LunarG just released new SDKs for Vulkan that include the most recent extensions. Much of the documentation for this SDK can be found in the Getting Started Guide, located in the Documentation directory of the SDK and on the Vulkan SDK Download site.

Overview of new SDK features
        It’s now easier than ever to use the Vulkan Configurator in an existing development environment. The updated tool can be used to enable/disable specific layers without affecting unspecified layers, creating a more intuitive user experience.
        Since the release of SDK, the following new extensions have been added to the Vulkan spec and may be supported by a Vulkan device. To query extensions available to a given physical device, call vkEnumerateDeviceExtensionProperties() or contact the hardware vendor.
        Improved validation coverage and bug fixes

Please note that RenderDoc is no longer included in the Vulkan SDK.


SDK download:

3D-Tech News Around The Web / GPU-Z 2.17.0 released
« on: February 27, 2019, 12:59:20 PM »
- latest GPU-Z @ Geeks3D
- latest GPU-Z @ TPU

    Added support for NVIDIA GTX 1660 Ti, Titan RTX, RTX 2080 Mobile, RTX 2070 Mobile, RTX 2060 Mobile, Quadro RTX 4000, GTX 650 (GK106), Quadro P5200
    Added support for AMD Radeon VII, Radeon HD 8400E
    Added support for Intel Amber Lake GT2 (Core i7-8500Y)
    Added support for detection of Radeon Adrenalin 2019 version
    Simplified some sensor titles: "GPU Clock", "Memory Clock", "Shader Clock"
    Unified "Memory Used" / "Memory Usage" sensor titles to "Memory Used"
    Improved crash reporter to ask for problem description and optional email, so we can follow up
    Advanced Panel dropdown will show "Memory Timings" option only when memory timing details are available
    OpenCL "Max Packet Size" is now formatted as unsigned value
    Instead of empty value show "None" when no OpenCL Built-in Kernels supported
    Added support for "None" in OpenCL DP, SP, Half FP Capability
    Fixed "File creation failed" message during GPU-Z startup
    Fixed GPU and memory load monitoring on RX 580 2048 SP
    Fixed missing default boost clock on GTX 1660 Ti and some Pascal cards
    Fixed missing fan sensors on RTX cards with no monitor connected
    Fixed crash on startup (Windows XP)
    Fixed crash when opening DXVA 2.0 report (Windows XP)
    Fixed power limit reporting on older NVIDIA cards
    Fixed crash during BIOS save on older NVIDIA cards
    Fixed incorrect VRAM reporting on 16 GB Vega
    Fixed various crashes caused by physical memory access


GPU-Z 2.17.0 + GeForce GTX 1080

GeeXLab - english forum / Checkerboard in GLSL
« on: February 26, 2019, 06:15:09 PM »
Very simple demo that draws a checkerboard pattern in GLSL.


- GeeXLab

GeeXLab - checkerboard in GLSL

DXVK is a Vulkan-based Direct3D 10/11 implementation for Linux which allows running 3D applications using Wine.

DXVK 1.0 has been released and here are the main improvements:

- Added DXVK_HUD=api option to show the D3D feature level used by the application. Does not work correctly for D3D10 at the moment (#705)
- The device filter now matches substrings rather than the full device name, which may help if the device name changes after driver updates on RADV (PR #921)
- Minor performance improvement on RADV by generating better shader code.
- Minor improvements to CPU overhead in some scenarios.
- Potentially improved frame time consistency on Nvidia in some games
- If available, the VK_EXT_memory_priority and VK_EXT_memory_budget extensions are now used to improve behaviour under memory pressure and to report available VRAM to applications more accurately, respectively.
- The state cache directory will now be created if it does not exist yet (#907)
- Resident Evil 2: Enabled new d3d11.relaxedBarriers option to improve performance by up to 10% in GPU-limited scenarios. Note that enabling this option for other games can cause issues.


Khronos Group Begins Work on a New Standards Initiative to Bring Vulkan GPU Acceleration to Safety Critical Industries

February 25, 2019 — Embedded World Conference —Nuremberg‎, Germany — The Khronos® Group, an open consortium of leading hardware and software companies creating advanced acceleration standards, today announces the creation of the Vulkan® Safety Critical Working Group, to enable safety critical industries to take advantage of advancements in GPU graphics and compute acceleration, and display control, at the highest levels of safety integrity.

Safety critical graphics are a key component for industries such as automotive, avionics, medical and energy. As display requirements become more advanced, safety critical graphics APIs must evolve to meet the industries’ needs. Khronos has led the industry in developing graphics API standards for safety critical environments via its OpenGL® SC Working Group, which released the OpenGL SC 1.0 specification in 2005,and the OpenGL SC 2.0 specification in 2016. Now, automotive and other industries are seeking advanced GPU graphics, compute and display functionality that can be deployed in safety critical systems.

In response to this industry demand, this new Khronos Working Group will create open, royalty-free API standards based on the existing Vulkan API specification to enable safety critical industries to utilize advanced graphics and compute acceleration. To best suit these market’s needs, the new API will aim to be compatible with industry standards for safety critical software, such as RTCA DO-178C Level A / EASA ED-12C Level A (avionics); FACE (Future Airborne Capability Environment) (avionics); and ISO 26262 ASIL D (automotive).

“Advanced graphics and compute functionality is being deployed into an increasing number of markets where safety is paramount, and so APIs designed to streamline the safety certification process are now a key part of Khronos’ standardization activities. OpenGL SC has been widely used in avionics systems, but now the wider embedded industry is demanding access to the latest GPU functionality. Vulkan is the ideal starting point for a new generation safety critical GPU API, as its driver architecture is significantly more streamlined than OpenGL, as well as offering increased control of device scheduling, synchronization, and resource management,” said Neil Trevett, president of Khronos.

Complete press release:

More about Vulkan SC

Vulkan SC

3D-Tech News Around The Web / Intel Graphics Driver released
« on: February 26, 2019, 01:09:16 PM »
- Latest Intel DCH graphics driver win10 64-bit @ Geeks3D
- Intel v6577 win10 64-bit @ Intel

Launch Driver
  - Dirt Rally 2.0
  - Sid Meier's Civilization VI: Gathering Storm
  - Apex Legends

- Performance improvements for Arma 3

- Enhanced support for DirectML

- Intermittent crashes or hangs may occur in Cinema4D, Sid Meier's Civilization VI: Gathering Storm*, vkQuake2
- Minor graphics anomalies may be observed in Final Fantasy XV* and Microsoft* Edge browser.
- OpenGL stability fixes for background applications.
- Few Internal panels may show black screen after installing graphics driver.
- Content protection may not enable when resuming from sleep after connecting display port monitor.
- Rotation may not be persistent when swapping displays in multi-monitor configurations.
- Aspect ratio options may be missing in Intel® Graphics Control Panel for HDMI displays
- Garbage may be seen while using Multi Stream Transport (MST) monitors.

Intel v6577 exposes Vulkan 1.1.97  but new device extensions (like VK_EXT_memory_budget or VK_EXT_validation_features) are not present...

Intel graphics driver v6577 information + UHD 630

Intel graphics driver v6577 + UHD 630 + GL-Z (OpenGL + Vulkan)

Pages: [1] 2 3 ... 55