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 - Stefan

Pages: 1 ... 101 102 [103] 104 105 ... 145
2041
SPIN-X created a little test [4.44MB] for the 3D game engine Spin-X Engine to test if graphics & physics engines run properly on different PC's.


Mushrooms are falling and the mushroom in the center bullies them out by spinning around clockwise.

Code: [Select]
Checksum: 0x5e9017ea
Operating System Information:
  Name: Microsoft Windows Vista Home Premium 64-Bit (6.0.6002)
  Service Pack: Service Pack 2
  Total Physical Memory: 4094MB (2677MB free)
  Total Virtual Memory: 8065MB (6503MB free)
CPU Information:
  Name: Intel(R) Core(TM)2 Quad CPU Q9450 @ 2.66GHz
  Description: Intel64 Family 6 Model 23 Stepping 7
  Manufacturer: GenuineIntel
  Architecture: x64 (64-bit) running on a 64-bit OS
  Number of Processors: 1 (4 Cores & 4 HW Threads)
  L2 Cache Size: 12288KB
  L3 Cache Size: 0KB
Memory Module Information:
  Bank0/1: 2048MB
  Bank4/5: 2048MB
Video Adapter Information:
  Name: NVIDIA GeForce GTX 465
  Total Memory: 1024MB
  Drivers: nvd3dumx.dll,nvwgf2umx.dll,nvwgf2umx.dll,nvd3dum,nvwgf2um,nvwgf2um
  Driver Version: 8.17.12.5819 (6/7/2010)
Monitor Information:
  Description: ACER AL2016W
  Native Resolution: 1680 x 1050
=========================================================
Loaded Direct3D9 (d3d9.dll v6.0.6002.18005)
Created windowed Direct3D9 device:
  Adapter: NVIDIA GeForce GTX 465
  Driver Version: 8.17.12.5819
  Resolution: 1280 x 720 [X8R8G8B8]
  Supported Shader Models: vs3.0, ps3.0
  Max Texture Resolution: 8192 x 8192
  Max Simultaneous Textures: 8
  Max Simultaneous RTs: 4
  Max Anisotropic Filtering: 16
Loaded NVIDIA PhysX (PhysXLoader.dll v2.8.3.31)
Initialized PhysX SDK (v2.8.1)
Created hardware accelerated PhysX scene
Average FPS: 124


2042
Quote
This article provides information about the preview feature for generating and loading binary OpenCL™ kernels in the ATI Stream SDK v2.1.
...
Related Resources

    * ATI Stream SDK v2.2 Product Page
    * Sample code: clbinary.zip

Misc. headlines at ATI already read "SDK v2.2" - hasty webmaster or lot of typos  ???

2043
3D-Tech News Around The Web / NVIDIA driver round-up
« on: July 23, 2010, 02:50:05 PM »
http://www.nvidia.com/object/tesla-winserv2008-64bit-198.17-whql-driver.html
http://www.nvidia.com/object/tesla-winserv2008r2-64bit-198.17-whql-driver.html

Quote
New in Release 197/198
Added support for CUDA 3.0
Changes and Fixed issues in Version 198.17
Added support for Tesla M2050.
Fixed an issue where the nvsmi utility would fail to run. 
Changes and Fixed issues in Version 197.03
This driver resolves fan speed issues reported with version 196.28 drivers.
Changes and Fixed issues in Version 196.28
Fixed the TCC driver to work with non-administrator applications. 

You might have already seen 258.96 for Geforce/ION, Verde and Quadro, here's the rest:

http://www.nvidia.com/object/tesla-win7-winvista-64bit-258.96-whql-driver.html
http://www.nvidia.com/object/tesla-win7-winvista-32bit-258.96-whql-driver.html
http://www.nvidia.com/object/tesla-winxp-x64-258.96-whql-driver.html
http://www.nvidia.com/object/tesla-winxp-x32-258.96-whql-driver.html

FreeBSD, Solaris and Linux - misc. driver branches have been updated:

http://download.nvidia.com/solaris/96.43.18/NVIDIA-Solaris-x86-96.43.18.run
http://download.nvidia.com/solaris/173.14.27/NVIDIA-Solaris-x86-173.14.27.run
http://download.nvidia.com/freebsd/96.43.18/NVIDIA-FreeBSD-x86-96.43.18.tar.gz
http://download.nvidia.com/freebsd/71.86.14/NVIDIA-FreeBSD-x86-71.86.14.tar.gz
http://download.nvidia.com/freebsd/173.14.27/NVIDIA-FreeBSD-x86-173.14.27.tar.gz
http://download.nvidia.com/XFree86/Linux-x86/96.43.18/NVIDIA-Linux-x86-96.43.18-pkg0.run
http://download.nvidia.com/XFree86/Linux-x86/96.43.18/NVIDIA-Linux-x86-96.43.18-pkg1.run
http://download.nvidia.com/XFree86/Linux-x86/71.86.14/NVIDIA-Linux-x86-71.86.14-pkg0.run
http://download.nvidia.com/XFree86/Linux-x86/71.86.14/NVIDIA-Linux-x86-71.86.14-pkg1.run
http://download.nvidia.com/XFree86/Linux-x86/173.14.27/NVIDIA-Linux-x86-173.14.27-pkg0.run
http://download.nvidia.com/XFree86/Linux-x86/173.14.27/NVIDIA-Linux-x86-173.14.27-pkg1.run
http://download.nvidia.com/XFree86/Linux-x86_64/96.43.18/NVIDIA-Linux-x86_64-96.43.18-pkg0.run
http://download.nvidia.com/XFree86/Linux-x86_64/96.43.18/NVIDIA-Linux-x86_64-96.43.18-pkg1.run
http://download.nvidia.com/XFree86/Linux-x86_64/96.43.18/NVIDIA-Linux-x86_64-96.43.18-pkg2.run
http://download.nvidia.com/XFree86/Linux-x86_64/71.86.14/NVIDIA-Linux-x86_64-71.86.14-pkg0.run
http://download.nvidia.com/XFree86/Linux-x86_64/71.86.14/NVIDIA-Linux-x86_64-71.86.14-pkg1.run
http://download.nvidia.com/XFree86/Linux-x86_64/71.86.14/NVIDIA-Linux-x86_64-71.86.14-pkg2.run
http://download.nvidia.com/XFree86/Linux-x86_64/173.14.27/NVIDIA-Linux-x86_64-173.14.27-pkg0.run
http://download.nvidia.com/XFree86/Linux-x86_64/173.14.27/NVIDIA-Linux-x86_64-173.14.27-pkg1.run
http://download.nvidia.com/XFree86/Linux-x86_64/173.14.27/NVIDIA-Linux-x86_64-173.14.27-pkg2.run

There are also new drivers for OpenSUSE 11.1, 11.2 and 11.3, but i'm not crazy enough to copy and paste 100 hyperlinks :P

2044
Quote
The NVIDIA® Tesla™ C2050 / Tesla C2070 computing processor board is a PCI Express 2.0 full-height (4.376 inches by 9.75 inches) form factor computing add-in card based on the NVIDIA Tesla T20 graphics processing unit (GPU).
This board is targeted as high-performance computing (HPC) solution for PCI Express systems.
The Tesla C2050 / Tesla C2070 is capable of running 515 GFLOPs/sec of double precision processing performance. Tesla C2050 comes standard with 3 GB of GDDR5 memory at 144 GB/s bandwidth. Tesla C2070 comes standard with 6 GB of GDDR5 memory.
Both the Tesla C2050 and Tesla C2070 can be configured by the end user to enable or disable ECC or error correcting codes that can fix single-bit errors and report double-bit errors. Enabling ECC will cause some of the memory to be used for the ECC bits so the user available memory will decrease to 2.625 GB for Tesla C2050 and 5.25 GB for Tesla C2070.

Download full documentation here

2045
AMD offered an introductory tutorial to OpenCL™ that was held alongside the 2010 Symposium on Application Accelerators in High Performance Computing (SAAHPC’10).

Download presentation

2046
Fermi Tuning Guide Version 1.2

For CUDA Driver API applications, the use of 32-bit device code within a 64-bit host application is no longer recommended, as support for this mixed-bitness mode will be removed in CUDA 3.2.  (Note that the CUDA Runtime API already requires that the device code and host code have matching bitness.)

CUDA C Programming Guide Version 3.1.1

Removed from Sections 3.1.6 and 5.2.3 the paragraph about loading 32-bit
device code from 64-bit host code as this capability will no longer be supported
in the next toolkit release.
In Section 3.2.6.3, removed the reference to the canMapHostMemory property
and mentioned that all devices of compute capability greater than 1.0 now
support mapped page-locked host memory.
Mentioned in Section 3.2.7.1 that host   device memory copies of a memory
block of 64 KB or less are asynchronous.
Fixed the maximum size of a 3D texture reference for devices of compute
capability 2.0 (2048 instead of 4096) in Section G.1.
Updated the paragraph about __fdividef(x,y) in Section C.2.1 to clarify
behavior depending on compute capability and compilation flag.

2047
Quote
The popular foliage middleware SpeedTree's latest version, 5.2, has just launched, bringing the total species count of broadleaves, shrubs and other plants in its library to over 125.

The new edition contains a number of new features: First, it lets developers add rolling wind effects and individual leaf motion, like the random twitch of single leaves, that can be exported via point caching for cinematics and animations. It has added a world-building toolset to its SpeedTree Modeler utility, which it says will help developers fill large areas quickly by setting the necessary number of placement rules, such as by slope, water location or the locations of other trees. In the Modeler, developers can see their changes in real-time.

Version 5.2 also adds the ability to develop and grow branches, trunks and leaves from arbitrary meshes. Finally, developers can now freely convert SpeedTree models to their chosen size and units. At the same time,the company says it's reduced the tool's memory demands by 40 percent over version 5.1.

“Version 5.2 delivers options that have been requested by both our game and animation customers,” said IDV President Michael Sechrest. "These new features ensure an even faster, more efficient path to bringing precisely-modeled, naturally-animated trees quickly to vast virtual areas.

[via]

2048
3D-Tech News Around The Web / Blender 2.53 Beta Released
« on: July 22, 2010, 07:46:53 PM »
The Blender Foundation just released 2.53, the first ‘Beta’ of the Blender 2.5 series. While the excellent 2.50 release log  has been updated with 2.53 features, there’s no ‘changelog’ since the previous version available yet (this may follow later). What *is* clear is that lots of bugs have been fixed, there’s a new ‘addons’ system  in place and there’s been a major overhaul in the sculpting tools. Also remember that this is the version that ‘Sintel’ was produced with. Go ahead and download now, we’ll follow asap with more information!

[via]

2050
What's new in 1.1.1

After more than 17 million downloads of VLC 1.1.0 in one month,
this is a release focused on numerous fixes and small improvements:

    * Multiple libVLC improvements and fixes
    * Windows video output fixes
    * DxVA2 decoding on ATI GPU, with Catalyst 10.7
    * Multiple fixes for interfaces
    * Various decoders and demuxers fixes
    * Updated translations
    * Various crashes and errors fixed


P.S.
GPU accelerated post processing works now fine with my rig.

2051
Quote
Professional video editing demands enormous computational power, especially when
you have HD video firmly established as the de facto standard and the increasing
popularity of 4K and other higher-resolution formats. Simply manipulating high
resolution footage is a laborious task, even with the most powerful workstations; editors
apply changes and then wait patiently for renders to complete. Full quality previews of
complicated timelines are only possibly after a lengthy export and encode process. In the
past, users demanding more interactive control or those with high productivity
requirements had no choice but to spend tens of thousands of dollars on custom
hardware solutions.  
Adobe Premiere Pro CS5 takes a bold step forward with its new Mercury Playback
Engine by leveraging the massive parallel computational power of a Graphics Processor
unit (GPU). With a quantum leap in performance afforded by the GPU, HD video
editing of complex projects suddenly becomes interactive. Most high-end workstations
today feature powerful GPUs, brining to all Premiere users an editing experience
previously available only to professionals with the highest production budgets.
The Mercury Playback Engine was an ambitious project and required close collaboration
between Adobe and NVIDIA®, the GPU manufacturer. This whitepaper is intended for
those readers interested in the technical background of GPU computing, and specifically
how the parallel processing capability of NVIDIA GPUs provides such a remarkable
advancement for video processing. Although the technical detail presented herein is
certainly not required for effective use of the Mercury Playback Engine, it can serve to
better educate hardware configuration and purchasing decision and to satisfy the
technical curiosity of those hardware-savvy users.

Get whitepaper here

Read more at Icrontic

2052
Quote
CUDA by Example, written by two senior members of the CUDA software platform team, shows programmers how to employ this new technology. The authors introduce each area of CUDA development through working examples. After a concise introduction to the CUDA platform and architecture, as well as a quick-start guide to CUDA C, the book details the techniques and trade-offs associated with each key CUDA feature. You’ll discover when to use each CUDA C extension and how to write CUDA software that delivers truly outstanding performance.

2053
Kodu is a new visual programming language made specifically for creating games. It is designed to be accessible for children and enjoyable for anyone. The visual nature of the language allows for rapid design iteration using only an Xbox game controller for input (mouse/keyboard input is also supported).

2054
Quote
If you’re a Windows developer keen to start tapping into the GPU to get a big boost in your application performance, then today is a good day.

This morning we released NVIDIA Parallel Nsight 1.0 Standard, the industry’s first development environment for GPU-accelerated applications that work with Microsoft Visual Studio. Parallel Nsight has been in beta since January and if you were one of the 8000 beta developers who gave us feedback during this time, we’d like to say a big thank you for all your great input.

It’s through your support that Parallel Nsight 1.0 Standard is ready for download today here, and is available to all Visual Studio developers free of charge.

Parallel Nsight has a ton of great functionality, such as allowing developers to debug CUDA C/C++, or DirectCompute applications on the GPU using the same familiar tools and techniques as on the CPU. Parallel Nsight is also the premier environment for graphics development, delivering the DirectX 11 tools required by graphics developers to efficiently develop top game titles and visual computing applications.

Anton Kaplanyan, lead researcher at CryTek, the developer that brought you Crysis and Far Cry, had this to say: “Parallel Nsight is the first toolbox in the world that allows us to look under the hood of the GPU, and makes parallel debugging not only possible but pleasant, significantly accelerating DirectX 11 development.”

George Tang, vice president and general manager of the Video and Home entertainment group at ArcSoft said that “NVIDIA Parallel Nsight has become our daily development tool when working with our CUDA-based applications such as SimHD and H.264 encoder.”

As always, we want to hear your feedback so that we can continue to produce the highest quality developer tools. For this reason, if you are a member of the Parallel Nsight Beta Program, your premium support account will remain active and you can continue to give us feedback via the Parallel Nsight Development Community Forums. And if you're joining us at the upcoming GPU Technology Conference in San Jose, CA, we'll also be hosting in-depth Parallel Nsight tutorial sessions. Happy Developing!

[via]

Quote
With this release, we now offer two versions of Parallel Nsight:
- 1.0 Standard (no cost): Includes the CUDA C Debugger, Shader Debugger, and Graphics Inspector.
- 1.0 Professional (Release Candidate 2): Includes the Analyzer, and CUDA C data breakpoints.

Getting the build:
You can grab the build from the Parallel Nsight Developer Zone page.
Follow the instructions on the download page to download and then activate your build.

For a limited time, we are offering FREE time-limited licenses of Parallel Nsight 1.0 Professional (Release Candidate 2). More information on getting a free Pro license is available on the download page.

Features:

CUDA C Debugger:
* Source debugging on the GPU hardware of CUDA C code.
* Supports the CUDA 3.0 and 3.1 toolkits
* Use breakpoints and single stepping to debug your code
* Use Locals, Watches, and Memory window to inspect variables and data on your GPU while debugging
* Use conditional breakpoints to break on a particular block and thread.
* (Pro only) Use data breakpoints to stop on a memory write to a particular memory region.


Direct3D Shader Debugger:
* Supports debugging HLSL shaders in programs using Microsoft Direct3D 11, 10.1 and 10
* Pixel, vertex, geometery, hull and domain shaders are all supported.
* Use Locals, Watches, and Memory window to inspect data on your GPU while debugging
* Use conditional breakpoints to break on a particular pixel

Direct3D Graphics Inspector:
* Support for applications using Microsoft Direct3D 11, 10.1 and 11
* Capture a frame for inspection at any time.
* Scrub through all draw calls to see how your frame was built.
* Examine all textures, render targets and GPU pipeline state in the frame.
* Use pixel history to see all draw calls that wrote to a pixel location.
* Resume your live application at any time! No need for lengthy capture => analyze => re-run app workflows.
* Frame Profiler - get per-draw call hardware bottleneck information for the ultimate in rendering optimization.

Analyzer (Parallel Nsight Professional only):
* Support for GeForce GTX 470 and GTX 480
* The Trace Activity now supports the following events:
** CUDA 3.1 Driver API calls, kernel launches, and memory copies
** OpenCL 1.0 Platform API calls, Runtime API calls, and device commands
** Direct3D 11 API calls and Direct3D 9 Performance Markers
** OpenGL 4.0 API calls and WGL
** Cg 2.2 API calls

Supported Operating Systems:
* Windows 7 and Windows Vista operating systems are supported.

Supported GPUs:
* GeForce: 9 series and better, including ION
* Quadro: 3800, 4800, and 5800 series cards
* Tesla: C1060, S1070, C2050, M2050/M2070

[via]

2055
3D-Tech News Around The Web / Direct3D 11 Tessellation resources
« on: July 21, 2010, 04:12:13 AM »
The new hardware tessellation feature available on Direct3D 11 video cards has great potential, but using it effectively currently requires understanding higher-order surfaces as well as a myriad of performance implications.
In addition to the Windows DirectX Graphics documentation, here are a number of resources for learning more about using Direct3D 11 Tessellation.

2057
Quote
Bohemia Interactive is pleased to announce the release of the hotly anticipated demo for Arma 2: Operation Arrowhead, a massive standalone expansion pack to the critically acclaimed military simulator, Arma 2.
To download the Arma 2: Operation Arrowhead demo visit the Demo page.

After maximising all graphics settings, GPU memory usage exceeded 1000 MB  :P
I'd like to see a performance test GTX460/768MB vs. GTX460/1024MB with this game.


2058
Quote
The main problem with my first ABuffer implementation (cf. my previous post) was that a fixed maximum number of fragments per pixel has to be allocated at initialization time. With this approach, the size of the ABuffer can quickly become very large when the screen resolution and depth complexity of the scene increase.

Full story at Icare3D

2059
The ATI Catalyst™ Software quarterly newsletter is designed to inform you about significant software features that have been released in the past quarter and about significant upcoming software features and milestones.   
This newsletter also includes a selection of the latest press quotes, and other important highlights

Download here

[via]

2060
This tutorial series is aimed at developers trying to learn OpenCL  from the bottom up, with a focus on practicality (i.e. I’m still learning, I’m sharing what I’ve found to work). Learning by example works best for me so make sure to get the code! It can only help you to have a copy of the OpenCL specification handy, and it doesn’t hurt (too bad) to read it!

Also check out:
20 million particles in OpenCL on the GTX480

Pages: 1 ... 101 102 [103] 104 105 ... 145