Archive for the ‘Nvidia’ tag
Nvidia on Real-time Physics
Short video, covering PhysX Technology overview from Jen-Hsun Huang GTC 09 keynote, is available now on Nvidia YouTube channel (in 720p and 1080p versions).
Clip demonstrates realtime demos of SPH Fluid and multigrid Eulerian Smoke simulations.
Related materials: GTC PhysX demos screenshots pack and GTC APEX Wall Destruction demo
Nvidia Q&A roundup: PhysX will not support OpenCL or DirectCompute any time soon
Under “Would you like to Ask Nvidia A question?” iniative Nvidia employees have answered another portion of questions, this time solely related to PhysX:
#1 – How do you expect PhysX to compete in a DirectX 11/OpenCL world?
By Tom Petersen, Director of Technical Marketing: PhysX does not compete with OpenCL or DX11’s DirectCompute.
PhysX is an API and runtime that allows games and game engines to model the physics in a game. Think of PhysX as a layer above OpenCL or DirectCompute, which in contrast are very generic and low level interfaces that enable GPU-accelerated computation. Game developers don’t create content in OpenCL or DirectCompute. Instead they author in toolsets (some of which are provided by NVIDIA) that allow them to be creative quickly. Once they have good content they “compile” a specific platform (PC, Wii, Xbox, PS3, etc) using another tool flow.
During this process game studios have three basic concerns:
1. Does PhysX make it easier to develop games for all platforms – including consoles?
2. Does PhysX make it easier to have kick ass effects in my game?
3. Will NVIDIA support my efforts to integrate this technology?
And the answer to the three questions above is: yes, yes, and yes. We are spending our time and money pursuing those goals to support developers, and right now the developer community is not telling us that OpenCL or DirectCompute support are required.
At the end of the day, the success of PhysX as a technology will depend on how easy it is for game designers to use and how incredible the game effects are that they create. Batman: Arkham Asylum is a good example of the type of effects we can achieve with PhysX running on NVIDIA GPUs, and we are working to make the next round of games even more compelling. At this time, NVIDIA has no plan to move from CUDA to either OpenCL or DirectCompute as the implementation engine for GPU acceleration. Instead we are working to support developers and implement killer effects.
#2 – Will PhysX become open-source?
Tom Petersen: NVIDIA is investing a lot of time and effort in PhysX and we do not plan to make it open source today. Of course the binaries for the SDK are distributed for free, and source code is available for licensing if game designers need it.
Only partial version is displayed. You can view full answer here
EVGA GTX275 CO-OP PhysX edition announced
EVGA GTX 275 CO-OP PhysX edition graphics solution was announced recently. As predicted, mysterious card turns out to be GTX 275 dedicated for graphics and GTS 250 dedicated for PhysX on single PCB.

Technical Specifications
| Core Clock Speed | 633+738MHz |
| Processing Cores | 240+128 |
| Memory Clock Speed | 2268+2200MHz |
| Memory Bandwidth | 127.0+52.8GB/sec |
| Shader Clock Speed | 1296+1836MHz |
| Bus | PCI-E 2.0 |
| Interface | DVI-I, DVI-I |

Recommended price is 349$.
Update: some pictures from Launch Event
Source: EVGA
ATI HD 5870 + NV GT 220 PhysX Benchmarks
Two questions may bother some people, interested in PhysX nowadays – “Are those hybrid ATI+NV PhysX configurations really working ?” and “How new GeForce GT 220 can perform as dedicated PhysX card ?“. User soothepain from coolaler.com forums has performed a bunch of tests, that can answer both of those questions.

System, used for benchmarks includes Intel Core i7 975 Extreme CPU, 6GB DDR3 RAM, EVGA Classified X58 Motherboard, ATI HD 5870 and Nvidia GeForce GT 220 GPUs running on Windows 7 RC7100 64-bit OS.
Three setups were tested:
Single GT 220 (Both graphics and PhysX processing)
Single HD 5870 (CPU is calculating all PhysX effects in this case)
HD 5870 for graphics + GT 220 for PhysX (using GenL Mod)
Test results:
Nvidia Halloween Launch Party: taking PhysX to the next level ?
Nvidia start to propagate invitation to Launch Party event, that will be hold at Nvidia campus in Santa Clara on October 30. New graphics card is going to be launched (no, not Fermi), with curious promise to “take PhysX to the next level”.

Insufficient details leaves some space for speculations, but we think that new GeForce GT240, another GPU in Nvidia’s 40nm line, will be revealed. Probably – new GPU PhysX title is going to be announced too.
Update: some sources indicates that it will be dual card with GT200b (graphics) + G92 (PhysX) GPUs, like GTX275 and GTS250 combined on one PCB.
And what’s your opinion ? tell us in comments.
You can sign to event here, and view full newsletter here.
Official Nvidia position on hybrid ATI+NV PhysX configurations

Back again on banned ATI+Nvidia PhysX configurations situation. Initially, all the hype was based on e-mail from customer care support, while Nvidia official motives were undisclosed. In order to bring some clarity, we’ve asked Nvidia for commentaries a while ago and finally recieved some answers:
PhysXInfo: Is it true that PhysX capabilities of NVIDIA GPUs and/or Ageia PPUs are disabled with newest drivers when a Non-NVIDIA (AMD) graphics card is present in the system?
Nvidia: It is true that running an NVIDIA GPU for PhysX with an AMD GPU for graphics is not a supported configuration. This mode was inadvertently turned on in a beta version of our Win7 drivers and on some older XP drivers. AMD does not support PhysX for their customers, and we don’t QA this configuration. With no QA, it is risky to run this configuration so we removed this capability in a recent driver release.
PhysXInfo: Can you please explain to us what factors have led to such decision?
Nvidia: Today NVIDIA’s GPU and PhysX drivers are interconnected to optimize performance. In the future we expect this interdependence to deepen. This alone makes it difficult to support a third party GPU.
In order to make sure our customers have a great experience, we QA every release of our PhysX or Graphics drivers by testing approximately 14 NVIDIA GPUs for graphics processing with 8 GPUs for PhysX processing on 6 common platforms with 6 OS’s using 6 combinations of CPU and memory. This is over 24000 possible configurations. While we don’t test every possible combination, it should be clear that the work and cost to NVIDIA is substantial. AMD does not support PhysX for their customers. Adding AMD GPUs would significantly increase the necessary work and cost for NVIDIA. We prefer to invest in inventing new technologies that give our customers great new experiences.
Reminder: Nvidia interview at WJPF will start in one hour
Interview with Nvidia about PhysX at WJPF news radio will start in one hour – at 1 PM CST.
Live stream is available at WJPF website (click “listen live” button at the right)
After the interview, Hi Tech Legion will provide more in-deph view into the PhysX technology.
Interview with Nvidia about PhysX live on WJPF news radio
According to High Tech Legion, WJPF news radio will be airing interview with Nvidia regarding PhysX technology at September 6, 2009, 1 PM CST (GMT-6).
Live stream will be accessible via WJPF website
Additionally, recorded interview will be available for HTL readers, along with some extra materials.
Source: Hi Tech Legion












