Lara Croft Joins GeForce with 45% Performance Gain
Release Summary
Improves performance and stability in Tomb Raider
Adds new SLI and 3D Vision profiles.
Feature Highlights of the AMD Catalyst™ 13.3 Beta Driver for Windows®
Improve performance in Sim City 5 up to 16 percent
Improves performance in Far Cry 3 up to 5 percent
Improves latency performance issues seen in Tomb Raider and Hitman Absolution
Resolves slight corruption seen in Tomb Raider with TressFX enabled for CrossFire and single GPU configurations
Made progress in Nexus 4 phone for using OpenCL in java & also made it to work with aparapi framework , so one can write pure java applications (using GPGPU & OpenCL) on the Nexus 4 devices. IMHO same approach can be used for any other future Android phones with proper OpenCL.so libraries deployed in the phones (without HW/SW vendor's SDKs).
Please find code at google-code
I/ActivityManager(471): START u0 {act=android.intent.action.MAIN cat=[android.intent.category.LAUNCHER] flg=0x10200000 cmp=org.aopencl.clinfo/.MainActivity} from pid 965
I/ActivityManager(471): Start proc org.aopencl.clinfo for activity org.aopencl.clinfo/.MainActivity: pid=3007 uid=10127 gids={50127, 1028}
D/libEGL(3007): loaded /system/lib/egl/libEGL_adreno200.so
I/CLInfo(3007): CLInfo:QUALCOMM Snapdragon(TM)
I/CLInfo(3007): 0:QUALCOMM Adreno(TM)
I/CLInfo(3007): Max Compute Units: 4
I/CLInfo(3007): Max work group size: 256
I/CLInfo(3007): Max Work Item Dimensions: 3
I/CLInfo(3007): Max work item dimensions: 3
I/CLInfo(3007): Max work items: (256,256,256)
I/CLInfo(3007): Preferred vector width char: 1
I/CLInfo(3007): Preferred vector width short: 1
I/CLInfo(3007): Preferred vector width int: 1
I/CLInfo(3007): Preferred vector width long: 1
I/CLInfo(3007): Preferred vector width float: 1
I/CLInfo(3007): Preferred vector width double: 0
I/CLInfo(3007): Max clock frequency: 325MHz
I/CLInfo(3007): Address bits: 32bits
I/CLInfo(3007): Max memory allocation: 981811200 bytes
I/CLInfo(3007): Image support: True
I/CLInfo(3007): Max size of kernel argument: 256
I/CLInfo(3007): Alignment of base addres: 512 bits
I/CLInfo(3007): Minimum alignment for any datatype: 64 bytes
I/CLInfo(3007): Denorms: False
I/CLInfo(3007): Quiet NaNs: False
I/CLInfo(3007): Round to nearest even: False
I/CLInfo(3007): Round to zero: False
I/CLInfo(3007): Round to +ve and infinity: False
I/CLInfo(3007): IEEE754-2008 fused multiply-add: False
I/CLInfo(3007): Cache type err:0
I/CLInfo(3007): Cache line size: 16 bytes
I/CLInfo(3007): Cache size: 32768 bytes
I/CLInfo(3007): Global memory size: 1963622400 bytes
I/CLInfo(3007): Constant buffer size: 4096 bytes
I/CLInfo(3007): Max number of constant args: 8
I/CLInfo(3007): Local memory type err:0
I/CLInfo(3007): Local memory size: 8192 bytes
I/CLInfo(3007): Profiling timer resolution: 1000
I/CLInfo(3007): Device endianess: True
I/CLInfo(3007): Available: True
I/CLInfo(3007): Compiler available: False
I/CLInfo(3007): Execution capabilities:
I/CLInfo(3007): Execute OpenCL kernels: True
I/CLInfo(3007): Execute native kernels: True
I/CLInfo(3007): Queue properties:
I/CLInfo(3007): Out-of-Order: False
I/CLInfo(3007): Profiling: False
I/CLInfo(3007): Name: QUALCOMM Adreno(TM)
I/CLInfo(3007): Vendor: QUALCOMM
I/CLInfo(3007): Error:CL_DRIVER_VERSION:-30
I/CLInfo(3007): Profile: EMBEDDED_PROFILE
I/CLInfo(3007): Version: OpenCL 1.1 Adreno(TM) 320
I/CLInfo(3007): Extensions: cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_qcom_default_relaxed_math cl_qcom_perf_monitor
I/CLInfo(3007): 1:QUALCOMM Krait(TM)
I/CLInfo(3007): Max Compute Units: 4
I/CLInfo(3007): Max work group size: 1
I/CLInfo(3007): Max Work Item Dimensions: 3
I/CLInfo(3007): Max work item dimensions: 3
I/CLInfo(3007): Max work items: (1,1,1)
I/CLInfo(3007): Preferred vector width char: 16
I/CLInfo(3007): Preferred vector width short: 16
I/CLInfo(3007): Preferred vector width int: 16
I/CLInfo(3007): Preferred vector width long: 1
I/CLInfo(3007): Preferred vector width float: 16
I/CLInfo(3007): Preferred vector width double: 0
I/CLInfo(3007): Max clock frequency: 1500MHz
I/CLInfo(3007): Address bits: 32bits
I/CLInfo(3007): Max memory allocation: 981811200 bytes
I/CLInfo(3007): Image support: True
I/CLInfo(3007): Max size of kernel argument: 256
I/CLInfo(3007): Alignment of base addres: 1024 bits
I/CLInfo(3007): Minimum alignment for any datatype: 128 bytes
I/CLInfo(3007): Denorms: False
I/CLInfo(3007): Quiet NaNs: False
I/CLInfo(3007): Round to nearest even: False
I/CLInfo(3007): Round to zero: False
I/CLInfo(3007): Round to +ve and infinity: False
I/CLInfo(3007): IEEE754-2008 fused multiply-add: False
I/CLInfo(3007): Cache type err:0
I/CLInfo(3007): Cache line size: 64 bytes
I/CLInfo(3007): Cache size: 1048576 bytes
I/CLInfo(3007): Global memory size: 981811200 bytes
I/CLInfo(3007): Constant buffer size: 65536 bytes
I/CLInfo(3007): Max number of constant args: 8
I/CLInfo(3007): Local memory type err:0
I/CLInfo(3007): Local memory size: 32768 bytes
I/CLInfo(3007): Profiling timer resolution: 0
I/CLInfo(3007): Device endianess: True
I/CLInfo(3007): Available: True
I/CLInfo(3007): Compiler available: True
I/CLInfo(3007): Execution capabilities:
I/CLInfo(3007): Execute OpenCL kernels: False
I/CLInfo(3007): Execute native kernels: False
I/CLInfo(3007): Queue properties:
I/CLInfo(3007): Out-of-Order: False
I/CLInfo(3007): Profiling: False
I/CLInfo(3007): Name: QUALCOMM Krait(TM)
I/CLInfo(3007): Vendor: QUALCOMM
I/CLInfo(3007): Error:CL_DRIVER_VERSION:-30
I/CLInfo(3007): Profile: EMBEDDED_PROFILE
I/CLInfo(3007): Version: OpenCL 1.1 Krait(TM)
I/CLInfo(3007): Extensions: cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_byte_addressable_store cles_khr_int64 cl_khr_fp16
D/libEGL(3007): loaded /system/lib/egl/libGLESv1_CM_adreno200.so
D/libEGL(3007): loaded /system/lib/egl/libGLESv2_adreno200.so
I/Adreno200-EGL(3007): <eglInitialize:269>: EGL 1.4 QUALCOMM build: Nondeterministic AU_full_mako_PARTNER-ANDROID/JB-MR1-DEV_CL2961380_release_AU (CL2961380)
I/Adreno200-EGL(3007): Build Date: 12/10/12 Mon
I/Adreno200-EGL(3007): Local Branch:
I/Adreno200-EGL(3007): Remote Branch: m/partner-android/jb-mr1-dev
I/Adreno200-EGL(3007): Local Patches: NONE
I/Adreno200-EGL(3007): Reconstruct Branch: NOTHING
D/OpenGLRenderer(3007): Enabling debug mode 0
I/ActivityManager(471): Displayed org.aopencl.clinfo/.MainActivity: +478ms
The purpose of this article is to discuss how to incorporate GPU accelerated effects into a project without negatively impacting the overall performance and framerate of a game.
The key to making this work is to run all visual effects independently of the main game thread and rendering pipeline. While the concept is simple, as always, the devil is in the details and there are a lot of things to consider to make this work properly.
Kayla Platform
ARM Development Platform for CUDA® and OpenGL
Introducing the Kayla Platform for computing on the ARM architecture – where supercomputing meets mobile computing.
The Kayla platform is powered by an NVIDIA® Tegra® Quad-core ARM processor and a Kepler™ GPU to deliver the highest performance, highest efficiency for the next generation of CUDA and OpenGL application. Pre-installed with CUDA 5 and supporting OpenGL 4.3, it provides ARM applications development across the widest range of application types.
The object of this competition is to create a cool JavaScript "application" no larger than 1k. Starting out as a joke, the first version ended with a serious amount of submissions, prizes and quality.
Intel® Graphics Performance Analyzers (Intel® GPA) is a powerful, agile developer tool suite for analyzing and optimizing games, media, and other graphics-intensive applications. The product supports applications intended for the Windows* OS platforms or Intel® Atom™ based phones running the Android* OS. The toolset is a free download from the Intel GPA Home Page.
Available now, Intel GPA 2013 R1 adds these key features:
Significant enhancements have been made to the Geometry Viewer within the Intel GPA Frame Analyzer tool. The Geometry Viewer enables developers to view the input geometry at various shader stages (IA, HS/DS, VS, GS) for each draw call in a captured frame, and helps developers debug difficult performance bottlenecks or complex geometry rendering issues. More details on specific enhancements are included later in this article.
A new command-line tool, gpa_console_client, is now available for analyzing Android* OS workloads. With this tool you can automate many functions and features of the Intel GPA System Analyzer that were previously only available from the GUI.
Support has been added for new Intel® Atom™ based phones running the Android* OS, such as the Motorola RAZR I*. The full list of supported phones is available here.
A new tutorial, City Racer, has been added to demonstrate a "hands-on" approach to performance analysis and optimization. This tutorial presents a step-by-step guide to using Intel GPA for performance analysis, bottleneck identification, and rendering optimization of a D3D11 application. After installing Intel GPA, this tutorial with full source code can be found here: <install directory>\GPA\2013 R1\sdk\Samples\CityRacer
Intel GPA System Analyzer and Intel GPA System Analyzer HUD now support games and graphics applications using the DirectX* 11.1 API. Developers using the latest DirectX* API are now able to analyze and optimize their games for hardware platforms that support this API.
New version available 4.1.0 !
This version addresses some issues with Windows 8, and OpenGL 4.3 support for all the platforms.
;----------1302271821-12.101-130214a-153749C-EDG_Direct
DriverVer=02/14/2013, 12.101.0.0000
AMD6763.1 = "AMD Radeon E6460"
AMD6743.1 = "AMD Radeon E6760"
AMD9804.1 = "AMD Radeon HD 6250 Graphics"
AMD9805.1 = "AMD Radeon HD 6250 Graphics"
AMD9807.1 = "AMD Radeon HD 6290 Graphics"
AMD9803.1 = "AMD Radeon HD 6310 Graphics"
AMD9802.1 = "AMD Radeon HD 6320 Graphics"
AMD9806.1 = "AMD Radeon HD 6320 Graphics"
AMD6720.1 = "AMD Radeon HD 6900M Series"
AMD9994.1 = "AMD Radeon HD 7400G"
AMD99A4.1 = "AMD Radeon HD 7400G"
AMD9992.1 = "AMD Radeon HD 7420G"
AMD9993.1 = "AMD Radeon HD 7480D"
AMD990A.1 = "AMD Radeon HD 7500G"
AMD9919.1 = "AMD Radeon HD 7500G"
AMD9990.1 = "AMD Radeon HD 7520G"
AMD99A0.1 = "AMD Radeon HD 7520G"
AMD9991.1 = "AMD Radeon HD 7540D"
AMD9904.1 = "AMD Radeon HD 7560D"
AMD9908.1 = "AMD Radeon HD 7600G"
AMD9918.1 = "AMD Radeon HD 7600G"
AMD9907.1 = "AMD Radeon HD 7620G"
AMD9917.1 = "AMD Radeon HD 7620G"
AMD9903.1 = "AMD Radeon HD 7640G"
AMD9913.1 = "AMD Radeon HD 7640G"
AMD9901.1 = "AMD Radeon HD 7660D"
AMD9900.1 = "AMD Radeon HD 7660G"
AMD9910.1 = "AMD Radeon HD 7660G"
AMD6819.1 = "AMD Radeon HD 7800 Series"
AMD6800.1 = "AMD Radeon HD 7970M"
AMD68B8.1 = "ATI Radeon HD 5700 Series"
Selected as Editor's Choice in the iOS App Store, featured in *TegraZone* and rated 4/4 "Must Have" by critics, BLADESLINGER is a visually stunning, fast paced 3D action brawler developed exclusively for high-end mobile devices.
BEST MOBILE GRAPHICS YOU’VE EVER SEEN
Bladeslinger introduces graphics features never seen before on mobile devices. Multiple mind-blowing 3D environments and characters, dynamic lighting and shadows with no detail spared. As mobile games approach console quality, Bladeslinger pushes the bar further than ever before.
TEGRA EXCLUSIVE FEATURES ★
Dynamic and progressive Storm Effects system: rain, lens drops, more reflective puddles, lightning and 3D positional thunder.
ArionBench is a software tool that allows you to benchmark the computing performance of your machine.
ArionBench is based on Arion 2 Technology (RCSDK), so it is capable of putting your CPUs and/or GPUs under heavy stress, to see how fast they perform when they must face the task of simulating the flow of light in a 3D scene.
ArionBench, and the list of results that we keep on this page, are fantastic tools for generic hardware analysis. But moreover, they are particularly valuable for Arion users who need to make a decision as to what CPUs or GPUs to buy for optimal rendering performance.
Requirements
An x86 computer and (recommended) an nVidia CUDA video card
ArionBench can run on any modern 32-bit/64-bit computer. However, to benefit from hardware acceleration (GPU), a nVidia CUDA video card with sm_20 (or newer) compute capability is required. Any GeForce card of the 400 series (or newer) and the corresponding Tesla/Quadro models are supported.
The new family of advanced NVIDIA professional graphics is fueled by NVIDIA Kepler™—NVIDIA’s most powerful GPU architecture ever—delivering unprecedented performance and innovative capabilities to boost your success. Whether you’re creating revolutionary products, designing groundbreaking architecture, reviewing the fine details in a CT/MRI scan, or telling spectacularly vivid visual stories, NVIDIA professional solutions let you do it better and faster.
GPU-Z is a lightweight utility designed to give you all information about your video card and GPU.
Revision History
0.6.8
Added support for GeForce GTX Titan, GeForce 505 (GT216), GT 640M, GTX 675MX, Quadro K2000, K3000
Added support for AMD Radeon HD 7480D, FirePro W8000, V8700, fake HD 6850 (based on HD 6450)
Fixed GT 620 GPU name
Fixed release date for Trinity HD 7000G
Added detection for Intel GMA
Improved GPU load monitoring on AMD cards
Fixed two application crashes related to Intel cards
The RuneScape Bestiary (free players and members)
This week, we're really excited to give you a glimpse at the next generation of RuneScape. We've created a new feature for the RuneScape website which lets you browse almost 3,000 of RuneScape's monsters and NPCs: the RuneScape Bestiary.
Each one is viewable in full 3D using our new HTML5 graphics engine - RuneTek 7 - which you'll be seeing much more of once we launch it for RuneScape later in the year. You can rotate, zoom and animate all of the available characters, share your favourites on Facebook, search for specific ones and view a variety of information about each of them
Release Highlights 313.26
Added support for the following GPUs:
GeForce GTX TITAN
Release Highlights 304.84
Fixed a bug that could lead to rendering corruption after an X server generation (i.e., leaving an X server running after all of its clients have exited).
Removed a spurious dependency on libpangox from the nvidia-settings binary shipped as part of the driver package.
Fixed a bug that could cause the X server to crash when performing an RandR 1.0 rotation (e.g., `xrandr --orientation left`) after unplugging the last connected monitor.
Fixed a bug that caused the nvidia-xconfig(1) man page to be blank.
Added support for X.org xserver ABI 14 (xorg-server 1.14).
Fixed font rendering performance and corruption problems on X servers with backported support for the new glyph cache functionality added to Pixman 0.27.
Fixed a potential X server crash during initialization, when a graphics card with a TV connector has no TV connected.
Added a new X configuration option, "UseHotplugEvents", to allow the suppression of RandR events when adding or removing non-DisplayPort displays. See the "X Config Options" appendix of the README for details
OpenCL Device Test
Lists the OpenCL processing characteristics of the connected OpenCL-compliant device. This includes the device's memory sizes, work-item/work-group sizes, and image processing support. This is only functional for devices with ARM's Mali GPU (the Nexus 10)
MSI Afterburner 3.0.0 Beta 6(2013-03-01)
There are no new features in this build. Only the following fixes:
- Fixed abnormal RTSS.exe CPU usage
- Fixed bug in thermal formula interpretator on triple temperature monitoring capable graphics cards (e.g. MSI R7970 Lightning), which could prevent application from starting properly on such cards
- Reverted to 32-bit time_t usage to provide backward compatibility with third party applications (gadgets and so on)
- Updated localizations
Stir is an interactive fluid simulation. Swipe the screen to stir up vortices, and be mesmerized by the ensuing perpetual swirly motion. To slow things down, press and hold the screen to dampen the whirlpools.
Choose from a number of presets that change the simulation behaviour and rendering style, or modify the parameters individually. Some notable parameters:
'Viscosity' controls how thick the fluid is. High values evoke materials such as molasses or paint. Low values are more lively, such as ink in water. Note that viscosity goes all the way to zero, allowing perpetual swirly motion! This is in contrast to many current fluid simulations that dampen and slow quickly.
'Symmetric' constrains the fluid motion to be symmetric about the centre of the screen. This leads to kaleidoscopic flows or perhaps the motion of deep space galaxies.
The underlying physical fluid simulation is based on novel research which support the two aforementioned features. Resource consumption(CPU, memory) is relatively lightweight and should work well on a range of devices. Of course the harder you push it (a key parameter being the number and size of particles) will determine what's possible, but even modest hardware should be adequate for lower particle counts.
This is a beta release. Any feedback including bug reports and development suggestions are welcome.
