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 4 ... 60
Geeks3D's GPU Tools / GPU Shark released
« on: July 04, 2019, 11:25:36 AM »
A new update of GPU Shark is available.

Download and release notes:

GPU Shark + GeForce RTX 2070

Terminus is a highly configurable terminal emulator for Windows, macOS and Linux

- Theming and color schemes
- Fully configurable shortcuts
- Split panes
- Remembers your tabs
- PowerShell (and PS Core), WSL, Git-Bash, Cygwin, Cmder and CMD support
- Integrated SSH client and connection manager
- Full Unicode support including double-width characters
- Doesn't choke on fast-flowing outputs
- Proper shell experience on Windows including tab completion (via Clink)


Terminus - terminal app

Terminus - terminal app

This is a proof of concept hair renderer that's based on a novel hybrid approach. It's capable of rendering strand-based hair geometry in real-time, even for fully simulated hair styles with over 100,000 strands. Our hybrid rendering pipeline scales in the performance/quality domain by using a rasterizer for closeup shots and a raymarcher for level of detail minification. We can render multiple hair styles at once with this hybrid technique, which has a smooth transition between the two solutions.


It is written 100% from scratch in Vulkan and simulates light scattering, self-shadowing and transparency in real-time with GLSL shaders and compute. We handle anti-aliasing by using a fast line coverage method. Our volumetric approximation is derived from the original geometry with a fast voxelization scheme that is used for raymarching and "direct" shading on an isosurface. We also use this volume for ambient occlusion and for more precise self-shadowing, even in the rasterized case.


Real-Time Hybrid Hair Renderer in Vulkan

THREE.js-PathTracing-Renderer is a real-time path tracer with global illumination and progressive rendering, all on top of the Three.js WebGL framework.

- Demo:
- Github:

THREE.js-PathTracing-Renderer real-time path tracer demo

Today we released PIX 1906.14, available for download here.  This release adds GPU timings to the New Timing Capture preview and further improvements to DXIL shader debugging.


    New Timing Capture Improvements
        Support for GPU timing data (see below)
        Many other usability improvements, performance improvements and bug fixes

    New pixtool features:
        New ‘launch’ options: –remote, –force11on12, –ignore11on12.  Run “pixtool –help launch” for more information.
        Added ‘take-new-timing-capture’ command.  Run “pixtool –help take-new-timing-capture” for more information.

    GPU Capture Improvements
        New event list control – can now pin columns, sort without losing hierarchy as well as general performance improvements
        Fixed issue with extremely slow capture-time performance when using reserved resources (previously this was worked around by copying dxcompiler into the same directory as the title)
        Fixed various shader debugging issues

    PIX has a new icon!


3D-Tech News Around The Web / NVIDIA Announces CUDA-X HPC
« on: June 17, 2019, 07:53:15 PM »
From fluid dynamics and weather simulation, to computational chemistry and bioinformatics, HPC applications span across many domains. Developing these applications requires a robust programming environment with highly optimized domain specific libraries.

Announced today, CUDA-X HPC is a collection of libraries, tools, compilers and APIs that helps developers solve the world’s most challenging problems. Similar to CUDA-X AI announced at GTC Silicon Valley 2019, CUDA-X HPC is built on top of CUDA, NVIDIA’s parallel computing platform and programming model. CUDA-X HPC includes highly tuned kernels essential for high-performance computing (HPC). GPU-accelerated libraries for linear algebra, parallel algorithms, signal and image processing lay the foundation for compute-intensive applications in areas such as computational physics, chemistry, molecular dynamics, and seismic exploration.



Light Tracer is an experimental engine for GPU accelerated, photorealistic client-side rendering inside a Web browser.

Client-side Web renderer

Light Tracer is the world's first solution for GPU accelerated, photorealistic client-side rendering inside a Web browser. At the moment, only Google Chrome and Mozilla Firefox are supported.

Physically-correct engine

It is based on a physically-correct unbiased ray tracing engine working on any graphics card – both discrete GPUs like NVIDIA/AMD and integrated GPUs like Intel (with a lower performance, of course).

Fast and efficient

Highly optimized GPU path tracing engine allows for instant feedback in the viewport while navigating 3D scenes. Interactive modifications made possible by partial acceleration structure rebuilds.


Light Tracer: Experimental GPU Ray Tracer for Web

3D-Tech News Around The Web / The Tech Behind Successful Indie Games
« on: June 17, 2019, 07:34:35 PM »
Some interesting stats about the technology used (what kind of game engine) to create some successful or very popular indie games:

The Tech Behind Successful Indie Games

3D-Tech News Around The Web / NVIDIA GeForce RTX 20 SUPER Series
« on: June 13, 2019, 03:11:11 PM »
NVIDIA plans to release shortly (July?) a new set of graphics cards, all more or less derived from the current RTX 20 series.

The following new graphics cards are planned:
- RTX 2080 Ti SUPER (based on a new GPU )
- RTX 2080 SUPER based on a TU104-450 with 3072 shader cores (+128 shader cores or one extra SM compared to current RTX 2080)
- RTX 2070 Ti SUPER (based on a new GPU )
- RTX 2070 SUPER based on a TU104-410 with 2560 shader cores (+256 shader cores or two extra SMs compared to current RTX 2070)
- RTX 2060 SUPER based on a TU106-410 with 2176 shader cores (+256 shader cores or two extra SMs compared to current RTX 2060)



3D-Tech News Around The Web / Vulkan Dos and Don’ts
« on: June 06, 2019, 08:38:29 PM »
After its popular DX12 Do's And Don'ts, NVIDIA publishes a new Dos/Donts, this time, for Vulkan:

The increased performance potential of modern graphics APIs is coupled with a dramatically increased level of developer responsibility. Optimal use of Vulkan is not a trivial concept, especially in the context of a large engine, and information about how to maximize performance is still somewhat sparse. The following document is a compilation of wisdom from some of the Vulkan experts at NVIDIA. It is not exhaustive, and is expected to be augmented over time, but should be a useful stepping stone for developers looking to utilize Vulkan to its full potential.


3D-Tech News Around The Web / Google speed test
« on: June 06, 2019, 07:41:43 PM »
Google has published a speed test to check the quality of your internet connection with Stadia.

This speed test provides a general indication of the quality of your connection, but does not guarantee playability on Stadia. Your gaming experience is dependent on a variety of factors, such as your device, internet service provider, and other bandwidth-consuming activities like downloads, file shares, and streaming.

Speed test:

I did a test:
- direct connection: I got 370Mbps
Your connection is great.
Based on your current download speed of 370.426 Mbps, we expect you’ll have a high-performance gaming experience on Stadia.

- connection via Opera's VPN: I got 17Mpbs
Your connection is good.
Based on your current download speed of 17.029 Mbps, we expect you’ll be able to play in high definition on Stadia, but you might not get 4K resolution.

We show how to compute global illumination efficiently in scenes with dynamic objects and lighting. We extend classic irradiance probes to a compact encoding of the full irradiance field in a scene. First, we compute the dynamic irradiance field using an efficient GPU memory layout, geometric ray tracing, and appropriate sampling rates without down-sampling or filtering prohibitively-large spherical textures. Second, we devise a robust filtered irradiance query, using a novel visibility-aware moment-based interpolant. We experimentally validate performance and accuracy tradeoffs and show that our method of dynamic diffuse global illumination (DDGI) robustly lights scenes of varying geometric and radiometric complexity. For completeness, we demonstrate results with a state of the art glossy ray tracing term for sampling the full dynamic light field and include reference GLSL code.


Dynamic Diffuse Global Illumination with Ray-Traced Irradiance Fields

3D-Tech News Around The Web / AMD gamedev / Vulkan presentations
« on: June 06, 2019, 11:11:34 AM »
The job of our worldwide developer technology engineers team is to directly help game developers to optimize their games, but also to educate developers by giving talks at various conferences and other events throughout the year. Recently I presented at Digital Dragons in Kraków (Poland), my colleague Lou Kramer visited Vulkanised in Cambridge (UK) to give a talk there, and Dominik Baumeister presented at Nordic Game.

- Optimising a AAA Vulkan Title on Desktop
- Triangles Are Precious – Let’s Treat Them with Care
- Make Your Game Friendly for Graphics Debugging and Optimization


3D-Tech News Around The Web / Python 3.8.0b1 available
« on: June 06, 2019, 11:04:17 AM »
Python 3.8 is still in development. This release, 3.8.0b1 is the first of four planned beta release previews. Beta release previews are intended to give the wider community the opportunity to test new features and bug fixes and to prepare their projects to support the new feature release.

Major new features of the 3.8 series, compared to 3.7
ome of the new major new features and changes in Python 3.8 are:
- PEP 572, Assignment expressions
- PEP 570, Positional-only arguments
- PEP 587, Python Initialization Configuration (improved embedding)
- PEP 590, Vectorcall: a fast calling protocol for CPython
- PEP 578, Runtime audit hooks
- PEP 574, Pickle protocol 5 with out-of-band data
- Typing-related: PEP 591 (Final qualifier), PEP 586 (Literal types), and PEP 589 (TypedDict)
- Parallel filesystem cache for compiled bytecode
- Debug builds share ABI as release builds
- f-strings support a handy = specifier for debugging
- continue is now legal in finally: blocks
- on Windows, the default asyncio event loop is now ProactorEventLoop
- on macOS, the spawn start method is now used by default in multiprocessing
- multiprocessing can now use shared memory segments to avoid pickling costs between processes
- typed_ast is merged back to CPython
- LOAD_GLOBAL is now 40% faster
- pickle now uses Protocol 4 by default, improving performance


For devs who embed Python in their apps (like in GeeXLab or in Blender), this PEP (Python Enhancement Proposals) is interesting:
Add a new C API to configure the Python Initialization providing finer control on the whole configuration and better error reporting.
- PEP 587 -- Python Initialization Configuration

python programming language

Geeks3D's GPU Tools / MSI Kombustor 4 v0.8.0.0 released
« on: June 05, 2019, 09:20:22 AM »
After nearly one year without update, MSI Kombustor is back with a new version. This version adds support of recent NVIDIA and AMD GPUs, fixes bugs in the Vulkan renderer and brings two new stress tests / benchmarks. The first benchmark (MSI-01) is a pure graphics stress test (tessellation + DoF + PBR) while the second benchmark is based on PhysX 4 and allows to test either the CPU (multi-threaded simulation) or the GPU. Benchmark scores can now be submitted online.



the FurMark stress test (OpenGL or Vulkan)
MSI Kombustor - FurMark stress test

the PhysX 4 stress test (CPU or GPU)
MSI Kombustor - PhysX 4 stress test

the tessellation / DoF / PBR stress test
MSI Kombustor - PhysX 4 stress test

This post is a walkthrough of the vulnerability that we discovered that allows execution of arbitrary commands on a system with the NVIDIA GeForce Experience (GFE) prior to version 3.19 installed – CVE-2019-5678. The exploit can be achieved by convincing a victim to visit a crafted web site and make a few key presses. This is possible due to command injection which was discovered in a local “Web Helper” server which GFE launches on startup.


MWR Labs stated in this blog post that GFE starts a local API server which allows control over different aspects of GFE. When you change a setting in the GFE GUI interface, it is likely just making a call to this local API. Knowing this, I thought it may be worthwhile to look into the API to see if there was any interesting functionality. The server that is started by GFE is NodeJS Express and many of the JavaScript source files can be found in “C:\Program Files (x86)\NVIDIA Corporation\NvNode”.



With Windows Version 1903, we have added a new toggle in Graphics Settings for variable refresh rate. Variable refresh rate (VRR) is similar to NVIDIA’s G-SYNC and VESA DisplayPort Adaptive-Sync.

This new OS support is only to augment these experiences and does not replace them. You should continue to use G-SYNC / Adaptive-Sync normally. This toggle doesn’t override any of the settings you’ve already configured in the G-SYNC or Adaptive-Sync control panels.

This new toggle enables VRR support for DX11 full-screen games that did not support VRR natively, so these games can now benefit from your VRR hardware.


Windows 10 VRR setting

NVIDIA Nsight Graphics is a standalone developer tool that enables you to debug, profile, and export frames built with Direct3D (11, 12, DXR), Vulkan (1.1, NV Vulkan Ray Tracing Extension), OpenGL, OpenVR, and the Oculus SDK.

NVIDIA announces Nsight Graphics 2019.3! In this release, GPU Trace has been revamped with a new analysis mode, the Configurable Range Profiler is now the default view in the profiling activity, and the Acceleration Structure Viewer can now export acceleration structures for standalone viewing. For full details on the new capabilities in the 2019.3 release, see our full release notes!

- Nsight Graphics 2019.3 announcement
- Nsight Graphics 2019.3 @ youtube
- Nsight Graphics homepage @ NVIDIA
- Nsight Graphics DOWNLOAD @ NVIDIA

Nsight Graphics 2019.3

Pages: 1 [2] 3 4 ... 60