Archive for the ‘Events’ Category

EGPGV 2012

15. May 2012

If you were wondering what’s up with last week’s ‘Introducing Lunchbox’ post: There wasn’t one since I’ve been to the Eurographics Symposium on¬†Parallel Graphics and Visualization to present our paper “Parallel Rendering on Hybrid Multi-GPU Clusters”. This week I’m attending Eurographics, but I’ll try to post the fourth article in the Lunchbox series by Friday.

Our paper presented a collection and evaluation of optimizations for medium-sized GPU clusters which use Multi-GPU NUMA nodes. This type of architecture is quite important, since it provides a cost-effective configuration for parallel rendering, since the host and network infrastructure cost is amortized over multiple GPUs. During this paper we found a few surprising insights (<cough>glFinish</cough>) on what optimizations are actually important.

Enough talk: The most important parts are summarized in our presention. Enjoy!


Four years of Equalizer

4. March 2009

Happy Birthday!

Last Sunday, we’ve passed the fourth anniversary of ‘Project Equalizer’, as it was called back then.

The name stuck, although we are way past a project definition. Equalizer has become a feature-rich, generic framework for creating parallel and scalable OpenGL applications.

Since last year, the code has matured considerably, and a lot of new exiting features such as load-balancing and DPlex (alternate frame rendering) support. Furthermore, the community has grown a lot and there are a couple of new users out there.

Since last year, the Equalizer project and community feels much more ‘serious’, something which can’t be expressed by features alone. The upcoming BOF and activity on the mailing list is a good indicator for this.

I am looking forward to another year of interesting tasks around parallel programming and 3D graphics!

Equalizer BOF at Eurographics 2009

16. February 2009

Good news everybody, we will be holding an Equalizer Birds-Of-a-Feather meeting during Eurographics’09 – mark your calenders:

– Place: Eurographics 2009, TU Munich
– Date: March 31, 15:00-16:30
– Room: To be announced

Co-located with EG is the Eurographics Symposium on Parallel Graphics and Visualization, so there is yet another good reason to come to Munich!

We are still looking for presentations, please contact me by end of February if you are working with Equalizer and would like to present your work.

Coincidentally, there is also a new poll:

Equalizer in Action

15. October 2008

In the near future there are two events where you can get a live Equalizer demonstration.

The first one is IEEE Vis next week in Columbus, OHIO. Max from the University of Zurich will give demonstrations in the Interactive Demonstrations Lab.

The second one is Siggraph Asia in December, where Renato Pajarola is contributing to the Interactive Massive Model Rendering course.

Feel free to pass by and talk to us!

Parallel Rendering Videos – Direct-Send

7. August 2008

The newest video explains parallel direct-send compositing for DB compounds. It is closely related to the Direct Send Compositing for Parallel Sort-Last Rendering paper we presented at EGPGV’07.

Note: the stuttering in the video comes from the capture tool, not Equalizer. ūüė¶

Equalizer 0.5 is coming

14. April 2008

The Equalizer 0.5 release was planned for end of February, but then a couple of conferences got in between.

Now it is already April, and we are back to it. You can already get the programming guide, either as as a hardcopy or as a download, after filling out a short survey.

Have a look at the release notes to see what’s coming.

Stay tuned for the release!

MacResearch WWDC Article

17. July 2007

MacResearch is running a nice interview with me about the WWDC event.

WWDC setup

25. June 2007

Since gWHIZ asked so nicely, here is a quick summery of the WWDC setup (list price as of today):

  • ¬†9 MacPro, ATI x1900 XT, 4 GB, 2×2 2.66GHz: $31.023
  • 19 30-inch Cinema Display: $34.181
  • 18 Vesa Mounts: $522
  • Gigabit Switch and Cabling: ¬†I don’t know what was used, but it’s cheap nowadays
  • Custom-built plywood wall: priceless!

The networking was probably the weakest link in the setup. All machines were on the same network. Optimisation is possible by putting the machines into two networks, both connected to the same frontend machine. Another optimisation can be implemented in Chromium by having a binary tree to submit the GL command stream.

Some money can be saved on the render nodes, since I don’t believe all the RAM is needed. Likewise, profiling might show that slower processors are sufficient.¬†

Back from WWDC

18. June 2007

Amira running through Chromium on the WWDC07 Display Wall

Yesterday I got back from WWDC – what a week! In addition to the prepared demo of Google Earth, Amira and VMD, we have managed to get two additional applications to run on the display wall through Chromium: LigandScout (molecular modeling) and SOAP (GPS satellite analysis).

The display wall was gorgeous, and even bigger than expected. We had a total of 18 30-inch Cinema Displays, arranged in a 6×3 layout. The other hardware was also standard Apple hardware you’ll get at the store: nine MacPro’s with a ATI x1900 graphics card, gigabit ethernet and one MacPro as the master to run the applications.

Google Earth running through Chromium on the WWDC07 Display Wall

People walking by were stunned by the size of the wall, and all applications looked fantastic and showed a variety of use cases. Often people lost a sense of scale and assumed we use 23-inch displays ūüėČ

In the afternoons I’ve had some time to attend the conference, and heard quite a few interesting talks.

Drop me a note or a comment if you’ld like to see a specific application picture.

WWDC, here I come…

7. June 2007

The new Equalizer polygon renderer

The last couple of days I have been busy preparing the WWDC demo of Chromium and Equalizer. We’ve made excellent progress, and certified quite a few applications (Google Earth, VMD, Paraview and Amira) to run on the display wall using Chromium.

On the Equalizer side, there is a new, improved version of the polygonal renderer using a much better internal data structure (a 3D kd-tree) and per-vertex normals. I haven’t been able to test it thouroughly, but I’ll have some time before the show to try it out on the MacPro cluster.

Tomorrow I’ll be travelling to San Francisco, and I am looking forward to an exciting show and to see you all! The scheduled times for the demos are Tue-Fri, 9am-12am.